{"_id":"create-ai-memory","_rev":"44-9cd27faf5e66fc96fea3f6ce9e552c0d","name":"create-ai-memory","dist-tags":{"latest":"0.19.1"},"versions":{"0.1.0":{"name":"create-ai-memory","version":"0.1.0","keywords":["ai","memory","claude","codex","gemini","cursor","opencode","cli","agent","obsidian","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.1.0","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"c8a81696896e07d1145a105aea1f06c1384b8cb9","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.1.0.tgz","fileCount":17,"integrity":"sha512-dlUQCqQnQkMdCuZJunX5g67pUZMVG12UwiUUB7oQywcLz3Ipr/RdbCEQfZxDe5FPzHLqczhyo5jNHLjCN4BMkg==","signatures":[{"sig":"MEUCIGDrU/3tC2fO32KOGJdepKl3x4QY8pYJqobI+UMnZFBiAiEAv1+etG9w+zZW1x4bQ8jMwB5Fz4YmO0zvn+GcFXKH3EA=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"unpackedSize":52924},"type":"commonjs","engines":{"node":">=18"},"gitHead":"8ca1b79616ec5a246d1bbfc3f32a940cfd39cea0","_npmUser":{"name":"rambaarde","email":"ramchrist20@gmail.com"},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"11.8.0","description":"Persistent, agent-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode). One Markdown vault, any agent.","directories":{},"_nodeVersion":"24.13.0","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.1.0_1783963426657_0.15103661697497017","host":"s3://npm-registry-packages-npm-production"}},"0.1.1":{"name":"create-ai-memory","version":"0.1.1","keywords":["ai","memory","claude","codex","gemini","cursor","opencode","cli","agent","obsidian","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.1.1","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"7dd1f8cd878ffa72b7a02de60c739a4bd712739d","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.1.1.tgz","fileCount":17,"integrity":"sha512-5L6oVEGj487yNhubYSLP4OoSOT2oo/B/142BwB7xjcHLB2FhAUiV5PgUioEuVBydmQSSXEWbt2j0sz608rUZcg==","signatures":[{"sig":"MEYCIQCXAp24ghsdNAmLysqjOcT9bqm5rcW8YT95x8QHaIiJ7wIhAO96tIW4IKje5u3L0zU10wHpfsaoUvQTGPeBjJY/htmG","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"unpackedSize":60776},"type":"commonjs","engines":{"node":">=18"},"gitHead":"75dede790ec402382dcec7b6e43185094a4436a8","_npmUser":{"name":"rambaarde","email":"ramchrist20@gmail.com"},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"11.8.0","description":"Persistent, agent-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode). One Markdown vault, any agent.","directories":{},"_nodeVersion":"24.13.0","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.1.1_1783967167186_0.2220830401800451","host":"s3://npm-registry-packages-npm-production"}},"0.1.2":{"name":"create-ai-memory","version":"0.1.2","keywords":["ai","memory","claude","codex","gemini","cursor","opencode","cli","agent","obsidian","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.1.2","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"202f5d1e83bab6bb62cf256b2779a9af912ff458","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.1.2.tgz","fileCount":17,"integrity":"sha512-+hg8TXSbIGPG9cM/2fIaVlZ/X/A6+4Ixp3HEPk6GLB+0tHhU44tHrX/bSji1l96FkkBIkzt4iTvSRP8XGS/toA==","signatures":[{"sig":"MEUCIQDp8egWCd8nGpm8t3OIWqd6E1XE/fgCpWD3KMMoeVo7vQIgJ7bLyZT6zOamKHz5apCFfkJsELNgWIHz4/VnhWZBqdo=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"unpackedSize":60897},"type":"commonjs","engines":{"node":">=18"},"gitHead":"c6a2a29bfe1cc3420f827d3fbe2eae035dbdfd11","_npmUser":{"name":"rambaarde","email":"ramchrist20@gmail.com"},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"11.8.0","description":"Persistent, agent-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode). One Markdown vault, any agent.","directories":{},"_nodeVersion":"24.13.0","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.1.2_1783967782847_0.17244726753789807","host":"s3://npm-registry-packages-npm-production"}},"0.2.0":{"name":"create-ai-memory","version":"0.2.0","keywords":["ai","memory","claude","codex","gemini","cursor","opencode","cli","agent","obsidian","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.2.0","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"ce9f038bd86e7e65c5ce8dd438a6ade1f334fe0f","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.2.0.tgz","fileCount":17,"integrity":"sha512-1UM9bFQLRGKBMyOidfv+ldACw9rjIvfWQ0A2grJsWmOCEmQIQeTGr3rjqbNqoR0OX2zP7nLnUYQc/kBhuEUIYQ==","signatures":[{"sig":"MEYCIQDgG1KxbtEgiGLiJ74POQUJ+6HR6CIcx+G6pO7qnaHcCQIhALy/SW4/6qYiKpRRcIwIhhdbtyLzwI29RxxXR9cTsI9e","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"sig":"MEQCICaU3RSycraerLh58VYMUp1eSy7rA7NJIlBV3U2cnpdVAiBQ5rRu4Gid43DhW2PZlkl/jwK75QD3zUbNsTfaWKe3Fg==","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.2.0","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":63833},"type":"commonjs","engines":{"node":">=18"},"gitHead":"cd848bd7bc23d8f62503fe9fcbec2ac316c5f888","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"oidc:40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.0.2","description":"Persistent, agent-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode). One Markdown vault, any agent.","directories":{},"_nodeVersion":"22.23.2","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.2.0_1787849896066_0.29216738133007536","host":"s3://npm-registry-packages-npm-production"}},"0.2.1":{"name":"create-ai-memory","version":"0.2.1","keywords":["ai","memory","claude","codex","gemini","cursor","opencode","cli","agent","obsidian","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.2.1","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"0294d5502cf883e1f4aeae479fb509953514e8be","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.2.1.tgz","fileCount":17,"integrity":"sha512-RTlMYzFbMplkcdNbnyv/fFqC5BsbZwXv/I1Baqn6fS+/gnGa5mMJ/0AK/m9TcS5gJTNAcWQWcae44gdcsOu9/w==","signatures":[{"sig":"MEQCIDEhazQzxUTfNr8LISxb2HECKshEWGoK8BAbLHOPbOFIAiAIUTmxvXC8CvXJtfnj0sBAhqM0B+7xP0hgQ8bZ2LuFoQ==","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"sig":"MEYCIQDtu/QUsrq1gTL4yjObPWjZHVXpqO8XUpio0dJAcsNr4wIhAJGVETW41NrXxpSiRF2qm3YwMU5JH9CmOvIhVVd/devV","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.2.1","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":64245},"type":"commonjs","engines":{"node":">=18"},"gitHead":"75d995a10a04bcac16ae49ddf309d2a0d5ea4d42","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"oidc:40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.0.2","description":"Persistent, agent-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode). One Markdown vault, any agent.","directories":{},"_nodeVersion":"22.23.2","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.2.1_1787887869648_0.5138081855984655","host":"s3://npm-registry-packages-npm-production"}},"0.3.0":{"name":"create-ai-memory","version":"0.3.0","keywords":["ai","memory","claude","codex","gemini","cursor","opencode","cli","agent","obsidian","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.3.0","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"f41e9d62e04e920f867f4c553687fe6743a74833","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.3.0.tgz","fileCount":17,"integrity":"sha512-LHZgSUScoZhjIW0jRhqaVWcbtUlPx6/YFxPIqkGre8zX9m2ntFrOxkptA7wHySCoiC4PRnvFhGtWXp7NaHpAHg==","signatures":[{"sig":"MEUCIQC1lz4J9d/xaKThfqTjpwEw+fKb4X5ewv5Z/D5aX3b4mgIgHAUvqlI/XQxOJD87o0D358BZXcUcyMJJ/vc6C74b16c=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"sig":"MEUCIQDUmViXoDTLK2bWLpdb9JNYSHSBxOsPK/iUjK6yFRGaNgIgAJFU9yxHQGB8YeMkpKHaRb0uxvnC29ummfHU98ryjPQ=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.3.0","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":65651},"type":"commonjs","engines":{"node":">=18"},"gitHead":"c0225f701115f7399aca21eeb69390e3876a850a","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"oidc:40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.0.2","description":"Persistent, agent-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode). One Markdown vault, any agent.","directories":{},"_nodeVersion":"22.23.2","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.3.0_1787900955771_0.06769121577603432","host":"s3://npm-registry-packages-npm-production"}},"0.4.0":{"name":"create-ai-memory","version":"0.4.0","keywords":["ai","memory","claude","codex","gemini","cursor","opencode","cli","agent","obsidian","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.4.0","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"177cd8ef3256272094a0973ceeecb7437067e8a5","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.4.0.tgz","fileCount":17,"integrity":"sha512-/j9dGKoDEdywmKhzikVoaBU26Cga4ujVhh+bI8B0RGODPC8zwrdrvKke2v1knSYtxMoVe0RtgjsghHn4W9vmDA==","signatures":[{"sig":"MEUCIQDxF1ctyfy5QEPkJ4Sj1JrEVkFPpQJY4Nye6KO5d3qBNwIgdQ2FfsxLDd8RlqrZQNX21FyFOyWm4lYbc0TrTrwOs4c=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"sig":"MEUCIQC2bR5OgYhHTfE753v3FmXKmcD6IvSPKz5XFVW1hvhvXgIgKlpUuEzEvwc9ldhoAsUiuzZHJqS5IDFMTjLW0hPqSC8=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.4.0","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":67608},"type":"commonjs","engines":{"node":">=18"},"gitHead":"0127cd06a1c4d6c4e05146cea1a957ffecb518bf","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"oidc:40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.0.2","description":"Persistent, agent-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode). One Markdown vault, any agent.","directories":{},"_nodeVersion":"22.23.2","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.4.0_1787901886154_0.4534735481312322","host":"s3://npm-registry-packages-npm-production"}},"0.5.0":{"name":"create-ai-memory","version":"0.5.0","keywords":["ai","memory","claude","codex","gemini","cursor","opencode","cli","agent","obsidian","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.5.0","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"e0f2003af510c3ff6d74550189244de13f1cd02c","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.5.0.tgz","fileCount":17,"integrity":"sha512-kvaPGl3KciW8Vvw38vleFntrC/YkuTcT+8UV/F0lNBeo3er4vaIpVYy9jQWHFS4aJGwFtp2EaUQJS+Z+Gaqtlg==","signatures":[{"sig":"MEQCIDCiy1A3TtboBtgx5Lu9nFe/ZvZBcN1J0vX0HpazXuFNAiAhMovfmyfEfo9jXa6NkNjPNopw35cNZXh0TR6TW6K0WA==","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"sig":"MEQCIHxz+yaK7puX/LRzZu8PjQVzS7+zh8fv1U2rVD3pvtQUAiBL7oK1P243VxxZA7czJZEfE5XK5pDK67ysGFbBXrLDCA==","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.5.0","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":68850},"type":"commonjs","engines":{"node":">=18"},"gitHead":"4000d04d5969d1231a6dfd9e0422bd5b9c21f0d9","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"oidc:40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.0.2","description":"Persistent, agent-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode). One Markdown vault, any agent.","directories":{},"_nodeVersion":"22.23.2","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.5.0_1787905930088_0.8805676677703376","host":"s3://npm-registry-packages-npm-production"}},"0.6.0":{"name":"create-ai-memory","version":"0.6.0","keywords":["ai","memory","claude","codex","gemini","cursor","opencode","cli","agent","obsidian","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.6.0","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"87f6f5be34764aab87f776dfb8455c9b8a9e0431","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.6.0.tgz","fileCount":17,"integrity":"sha512-j+OXvs8o3TfYtuVal85+ZAmJBxXMe4hij0H8xT2OQHNTHMW4Rv+97K/RJOsGHl5PkdTbySkGJQebqJvp3NmlYg==","signatures":[{"sig":"MEQCIAVEeignB3pcRMzAtH70H0aL2/LF81RVSafSRz3+DnaWAiBMoGNP+hDIpacKR5n0LnOqQ8Fee49oFeR4Xd4ZWuEjLA==","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"sig":"MEUCICtcCZ/IvhXI00rNrvwF7Wg/jao2Ksu7qldbk2FHE/7sAiEAnVRgL4VQSaaE0eMene4Tz2lSUKYECHmDbjpnAYCklVk=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.6.0","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":71051},"type":"commonjs","engines":{"node":">=18"},"gitHead":"f279f6b6464b820b4dfefc944f13b03a9dcdb395","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"oidc:40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.0.2","description":"Persistent, agent-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode). One Markdown vault, any agent.","directories":{},"_nodeVersion":"22.23.2","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.6.0_1787952700640_0.3276427759484073","host":"s3://npm-registry-packages-npm-production"}},"0.7.1":{"name":"create-ai-memory","version":"0.7.1","keywords":["ai","memory","claude","codex","gemini","cursor","opencode","cli","agent","obsidian","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.7.1","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"2d7f058e02e48a7a4ac4a6d7eb0755ef7bf973b6","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.7.1.tgz","fileCount":17,"integrity":"sha512-3CjiNgNs9Qq/iQ3nhAJFhzMtoT2uhx10rzgg7IUUtjeCffQbFKySBL4mJ1fAHe4kvrverpWqXOqIdPC+ZqeFqw==","signatures":[{"sig":"MEUCIQCcMPlVl1WtEPkiKHSRr9MKRForvWG+3nG7BqOQxw4QZQIgPzGGFF+t/yTWTj6D794Q+W6JI6osDWRhR+IhaNHDULg=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"sig":"MEYCIQCgL70Fk2JtRCUEU38SY3iXwETV/wmdxaCT/S6cPBYy8gIhAI9vkp2n7wzrirJtTZHtfJE4r8brGF3uG36nPo58CPZN","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.7.1","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":72361},"type":"commonjs","engines":{"node":">=18"},"gitHead":"7c6a3dde751772e29daf2ef19584550b67bf6dad","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"oidc:40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.0.2","description":"Persistent, agent-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode). One Markdown vault, any agent.","directories":{},"_nodeVersion":"22.23.2","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.7.1_1787966348380_0.9389583759470836","host":"s3://npm-registry-packages-npm-production"}},"0.8.0":{"name":"create-ai-memory","version":"0.8.0","keywords":["ai","memory","claude","codex","gemini","cursor","opencode","cli","agent","obsidian","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.8.0","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"850a403bc72e92187e346eb53702aa03b14d166d","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.8.0.tgz","fileCount":18,"integrity":"sha512-PmvThFUl8/5Yl1/yDb51oMxg7q6P3zNOqWL0JxxQoQ37+eP9RWUaSVR3Zm8dT+oXWJa3nKldS1e6CTes3M3edA==","signatures":[{"sig":"MEQCIGaJSbT1m0bSI3Yw/cHWRC5AAclasKPofP5zciDx0FC/AiBBvz1+CUyiVYI4X1cQK3CLoVmJOXb/EnePR3dSpcfazw==","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"sig":"MEUCIEz0pFAbsqkSW78g0em18XX8CGMqCQL2hNnak4DhAKxFAiEApu8TL3aOLFygGgrHLtcfULZZb0Ejvu5U+qZOkADgJ5A=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.8.0","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":76199},"type":"commonjs","engines":{"node":">=18"},"gitHead":"f35efc0225dc709ed207662db23c3ec7637d6679","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"oidc:40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.0.2","description":"Persistent, agent-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode). One Markdown vault, any agent.","directories":{},"_nodeVersion":"22.23.2","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.8.0_1788066166000_0.07457765040843367","host":"s3://npm-registry-packages-npm-production"}},"0.8.1":{"name":"create-ai-memory","version":"0.8.1","keywords":["ai","memory","claude","codex","gemini","cursor","opencode","cli","agent","obsidian","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.8.1","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"33f28c44d2c376fdc95a08ca11ba0dd2d94f48f8","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.8.1.tgz","fileCount":19,"integrity":"sha512-F7HlWCd4F2yU7GjhzTme9qFkYx2nvZcoQsINz4GVnjHBenxhuiaQ65wreZ/bPoqqkkh8CwFf9prmVEQrxVhVqA==","signatures":[{"sig":"MEUCIA70IGjwhmv7qaYEMmkOib6m/e2ysBHM28AjAUbb1bGwAiEAjH1gSAZzPeLozNi3Eo8RtWUsIaL4wO2w+kMqxmaxksQ=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"sig":"MEUCIB7HWyagi1JUVM7s3qd/YWMt/h/quLxww6rQpu3qdJGoAiEAwdjscZeri/d2cZsztWSSQfbcnrrkYbj8RA/tVHIct60=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.8.1","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":76904},"type":"commonjs","engines":{"node":">=18"},"gitHead":"0c04342ff5e56a0d2acc9b6a7b0f509712cd3ce0","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"oidc:40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.0.2","description":"Persistent, agent-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode). One Markdown vault, any agent.","directories":{},"_nodeVersion":"22.23.2","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.8.1_1788073910235_0.5889171414944869","host":"s3://npm-registry-packages-npm-production"}},"0.8.2":{"name":"create-ai-memory","version":"0.8.2","keywords":["ai","memory","claude","codex","gemini","cursor","opencode","cli","agent","obsidian","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.8.2","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"8f03c561e68c3fed1a9fbf4595301319e8105690","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.8.2.tgz","fileCount":19,"integrity":"sha512-8byzN/VlhNzl86Sws/rPsZ5IFlRsMZHG+9OTAcCmMlsUYHv7nXZSqDi5tX95l0L3gVTE207QS213YEsyOwa8jg==","signatures":[{"sig":"MEUCIQDhvhryOLwMDNsKC+exSRqaCtk27S5xxBpREK/9nOo3uQIgTjCqvBl/4rbW532pWdMFZOap5tizCV4dRxJ0zGnHs3o=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"sig":"MEUCIQDuyFa0xZSahAL0p5WdpeJDnQLaWo27XMqUrwLdH240tgIgRchqMXhD1BlmKD77jGNzPDvuJHnks9t1GT0TiKZ2ZdY=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.8.2","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":76908},"type":"commonjs","engines":{"node":">=18"},"gitHead":"d79c702e3d598a3c860f03baf6ae8b712b4154b8","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"oidc:40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.0.2","description":"Persistent, agent-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode). One Markdown vault, any agent.","directories":{},"_nodeVersion":"22.23.2","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.8.2_1788074435430_0.7008145157664161","host":"s3://npm-registry-packages-npm-production"}},"0.8.3":{"name":"create-ai-memory","version":"0.8.3","keywords":["ai","memory","claude","codex","gemini","cursor","opencode","cli","agent","obsidian","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.8.3","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"ece83692f6757121bff136cf28c4c52d04dca17f","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.8.3.tgz","fileCount":19,"integrity":"sha512-zj3VYyq+GOCddV5dKmCRgM8mf0phsUpUzT8U9mfgL9f2oW4C/LsrgThfTaVOiFAjsfl1omZyy/7uV/Y1RKhEWw==","signatures":[{"sig":"MEUCIQCGs+lVqOx8gYGEqLqXn4jR4T5aqBLU9GSWwdpfFDgYvgIgDQCT8FfKOtOotXSY2vqs9gQljtbs2718e5RUqCFBlgU=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"sig":"MEQCICfiWEJZ77Q7Hb5NVH+ZiYa7PXd5N5nkNs5b9+pr6FiFAiAQ4H7Il8TLDPFCf6I+weJcvPYtPHMklJonXke9E23KRQ==","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.8.3","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":77122},"type":"commonjs","engines":{"node":">=18"},"gitHead":"b9ca809e009e8ea7bd345f89340457ad3a406e5f","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"oidc:40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.0.2","description":"Persistent, agent-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode). One Markdown vault, any agent.","directories":{},"_nodeVersion":"22.23.2","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.8.3_1788075143277_0.3715579885023048","host":"s3://npm-registry-packages-npm-production"}},"0.9.0":{"name":"create-ai-memory","version":"0.9.0","keywords":["ai","memory","claude","codex","gemini","cursor","opencode","cli","agent","obsidian","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.9.0","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"e9c09be13c97a02de43f89cb44ed6c489b3b8e0d","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.9.0.tgz","fileCount":19,"integrity":"sha512-IfjcyTKop7/2jLzPAVcmMsZprfSBFXmnW9YifRKd+iO8cJLDXJzNm2JpeR8qhS39foysNF96gLg5jr+WTGwu4Q==","signatures":[{"sig":"MEUCIQCecoNf+z+6+6dqc76u6X2DJ7Ldi2e3lXh0tz8PrJhpLgIgNgeYrywccj7j836ECG3jMRUZlCyOBJhrLVA/Pogfhvw=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"sig":"MEYCIQCNm1PASPbUJdKCXNw8LSziD5NySq5UUiFgNuX19wyA7gIhAKDxR/cKh8XLhCzCttUhJ4xr8T+n50pfV9QBe+d2lxQP","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.9.0","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":77057},"type":"commonjs","engines":{"node":">=18"},"gitHead":"d568b5898cdecdeb944a7e264eb55410a19fc860","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"oidc:40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.0.2","description":"Persistent, agent-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode). One Markdown vault, any agent.","directories":{},"_nodeVersion":"22.23.2","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.9.0_1788075907777_0.35867031982950226","host":"s3://npm-registry-packages-npm-production"}},"0.9.1":{"name":"create-ai-memory","version":"0.9.1","keywords":["ai","memory","claude","codex","gemini","cursor","opencode","cli","agent","obsidian","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.9.1","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"44dfbaa78b94c54ec9d485b704ede98f2c6963b2","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.9.1.tgz","fileCount":19,"integrity":"sha512-FDlhN5J5AWo1Zq4B9ckugqiNPH2DDMNQt/XKZYiZuiJJ0YLQVCl3UH2J2G3YlbdQiLbSmVM3BIstTmzRgljELA==","signatures":[{"sig":"MEUCIQCaGyBcbOsgVx86oji/UWgros10md+N+x2+Oq9ZTec6iAIgJqlgkyYLR8WL/ui0zjc/OJbP8YhchdUV5709dINIrsI=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"sig":"MEQCIE5gibQQBhp0RbfoEOleJwn1CrKOwo8kSDxqjzStUgMGAiAFebg2GErBR69X5uW1v9upFwFfb+WVCUfHtccIxf039g==","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.9.1","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":77192},"type":"commonjs","engines":{"node":">=18"},"gitHead":"7926a54bdaa18e3a6fe44a1711c880219521a9fe","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"oidc:40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.0.2","description":"Persistent, agent-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode). One Markdown vault, any agent.","directories":{},"_nodeVersion":"22.23.2","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.9.1_1788076462017_0.21177793676385215","host":"s3://npm-registry-packages-npm-production"}},"0.10.0":{"name":"create-ai-memory","version":"0.10.0","keywords":["ai","memory","claude","codex","gemini","cursor","opencode","cli","agent","obsidian","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.10.0","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"9dc524d83768b2206d46e83b98dcf80fbae6d6f0","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.10.0.tgz","fileCount":19,"integrity":"sha512-8/nGiOuJaZ/ujoWJ937wsOK3PF0tXmFoIOz+XLkSMHQ+nQNzAYh0JxHCEz5vhDxzjyOE10Rsz3w35/0QREc8Rw==","signatures":[{"sig":"MEUCICcyJAtBcD4UC+mB5v5l6quNAZHvRdAb/AEZpxkTpysxAiEA8+ADnsOufmoinvcdvHq3URduLKEpwxMHjuCD5Y0giY0=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"sig":"MEQCIDINl7266Y1aErll4MGrK2fh5O3RnP7xQbMd6XC8JTksAiAxh6dV/ZM7QNujxrpojTTWvXL8VjDVVuwZdj3flsMoow==","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.10.0","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":79997},"type":"commonjs","engines":{"node":">=18"},"gitHead":"e3fbbff9b03604ec6a77899263b7abe5d8a5ff5f","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"oidc:40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.0.2","description":"Persistent, agent-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode). One Markdown vault, any agent.","directories":{},"_nodeVersion":"22.23.2","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.10.0_1788095689814_0.8513064591430541","host":"s3://npm-registry-packages-npm-production"}},"0.11.0":{"name":"create-ai-memory","version":"0.11.0","keywords":["ai","memory","claude","codex","gemini","cursor","opencode","cli","agent","obsidian","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.11.0","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"de7940dbace229b654a5818a90ed415fc372b947","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.11.0.tgz","fileCount":19,"integrity":"sha512-PKoBvhG3SjyznrGHoRbFgxlVhg+PHRWzv4bCx90JH6Ru8boF2yzbRNISdUcf58DT4xo05YRAZzBM4WIStmTX9w==","signatures":[{"sig":"MEUCIDewI9VHmVnjJl6ijA6FqZGbndoEp1UwOqvOUXGFVuXwAiEArpV/RacSD9JO4cqhM0Z03kcKuqbdIA1QNcrFgpUE2P4=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"sig":"MEQCIDfvLJd/fTKTMBDqqy3IobCrtqnb7DsmgY9KAu/P+yZqAiBOMQSrzaGQkFPqaInLn8cWmswF+jTq+lDjRpB2uy5W5A==","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.11.0","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":81420},"type":"commonjs","engines":{"node":">=18"},"gitHead":"c95da7e38787fd49125fb0adb0ee9a7e91da3683","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"oidc:40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.0.2","description":"Persistent, agent-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode). One Markdown vault, any agent.","directories":{},"_nodeVersion":"22.23.2","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.11.0_1788096359339_0.5144727083758573","host":"s3://npm-registry-packages-npm-production"}},"0.11.1":{"name":"create-ai-memory","version":"0.11.1","keywords":["ai","memory","claude","codex","gemini","cursor","opencode","cli","agent","obsidian","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.11.1","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"425c97095c65c858fe93333b8131f922055a5d81","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.11.1.tgz","fileCount":19,"integrity":"sha512-BNyh+sr0Ne1sW/EH604Hg9+iH5Vl5KAeEOxYKlFz7RFsFtUWUZYON91EREeK7J0PD7vZpCKkTlj4lPh65xonoA==","signatures":[{"sig":"MEYCIQD9t77FJAeyU40X3rEK9cf5zLH1P5xrq14F/9YMuWvpywIhAIdo0xTOKsrFhBB+TxlgZlGbPM0FbWrAwzqEVQJe2GhH","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"sig":"MEQCIAX3wUlMA1QQ6f4i/i1VDrz7AOV8wGpTlyq9bLLdK+W1AiAf4sLeLDVRKlr/Sjf11JpmynLPiUDDoy9x9ZNSDg+hww==","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.11.1","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":81384},"type":"commonjs","engines":{"node":">=18"},"gitHead":"afd6cf90fb70266da9bf4b19459ef2bcf2c29e7d","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"oidc:40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.0.2","description":"Persistent, agent-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode). One Markdown vault, any agent.","directories":{},"_nodeVersion":"22.23.2","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.11.1_1788096880404_0.3040449321916212","host":"s3://npm-registry-packages-npm-production"}},"0.11.2":{"name":"create-ai-memory","version":"0.11.2","keywords":["ai","memory","claude","codex","gemini","cursor","opencode","cli","agent","obsidian","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.11.2","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"68f58df6269eff62862f87364a6f0110ffc904e6","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.11.2.tgz","fileCount":19,"integrity":"sha512-MK898sfEntkHNIxkw4fDhEt8gR05IjoZnuvvi0NUAXyg/wjpgxGg2yZ1wtUD/qdf52ddOihAMTjCe3dz4YgvDw==","signatures":[{"sig":"MEUCIQD8WYSlg7fzuevBC1WvMDViDb9ZRmYUTLCOOkfTJd4DtAIgLcF1IptVEA9Vipq/ZE4UBlO59ULQu3DRzZFJOfmaO7E=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"sig":"MEUCIQC3oPiqiYcCnfCguA3A7JtPYQJInh18MqA7dBDeuwOfsgIgQ/KP9BS/i7JoDod9hKEv2vM9qDoI1eizSsnot9Hkr+A=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.11.2","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":83884},"type":"commonjs","engines":{"node":">=18"},"gitHead":"96590d47ea46af2903fab3a21f33c3e8b3f90edb","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"oidc:40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.0.2","description":"Persistent, agent-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode). One Markdown vault, any agent.","directories":{},"_nodeVersion":"22.23.2","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.11.2_1788098075072_0.9517095348468954","host":"s3://npm-registry-packages-npm-production"}},"0.11.3":{"name":"create-ai-memory","version":"0.11.3","keywords":["ai","memory","claude","codex","gemini","cursor","opencode","cli","agent","obsidian","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.11.3","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"7092aa522bf801ec3a5195963a1b29b203f1e379","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.11.3.tgz","fileCount":19,"integrity":"sha512-SSU62GDYBRC907yJM6W3j8tE5NhkPSESSGINWEIWvLS2berjhJSm1kOo4SRZMgCBVwRr4xf0y5U4fu5HWeXLgA==","signatures":[{"sig":"MEYCIQC15rLofNyyl7Tvko6ig9iB0hJPnzT3IFwVX/qXocGnkgIhAJs24JFtD11z7ExnNOOP3P4cn9Jc43VoHBKlEQm6MCxk","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"sig":"MEYCIQCHEXc7Dwq4P9wYEEHyG7Xar4fWfUflCe4+ZlGRTrUD0wIhAMVpl6hRiIo/DWX4xUhJkH9jp1mjZpDpHApwTPmK2oaZ","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.11.3","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":84316},"type":"commonjs","engines":{"node":">=18"},"gitHead":"7c754890bfe24b6f6bb15c4dacbdecea3853f79e","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"oidc:40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.0.2","description":"Persistent, agent-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode). One Markdown vault, any agent.","directories":{},"_nodeVersion":"22.23.2","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.11.3_1788098582440_0.3349116268099994","host":"s3://npm-registry-packages-npm-production"}},"0.12.0":{"name":"create-ai-memory","version":"0.12.0","keywords":["ai","memory","claude","codex","gemini","cursor","opencode","cli","agent","obsidian","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.12.0","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"b350cc61570307b90f46a130dd4e259983cdec85","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.12.0.tgz","fileCount":19,"integrity":"sha512-P7TdsgZv+/vAqji+xGSjJ88Pg/Qde0ISCQKjc5fAYr64PthhP843OEfWLq0DMU5+Fx5hjJBpqEvI6hl92Z5u7g==","signatures":[{"sig":"MEUCICuZuEdgeFrVuvpBgKxCL4DjXjXqu80i1BHjMk8bxKFkAiEA1hR7jvt08n7ksHT3C1aQSizY1y5t9ipG/8ncpKwNVdo=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"sig":"MEQCIDvfGZgNCW4uyNfTKv02B7PiMDb6DTtjsV1V3+VujEpGAiB1APA16lR7SrQ+6zZCZ6LTtlxsPFgyRYYRVfNVrlZvRA==","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.12.0","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":85705},"type":"commonjs","engines":{"node":">=18"},"gitHead":"f69a41643672bb17dd49e8dec2ec5daabc0a7860","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"oidc:40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.0.2","description":"Persistent, agent-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode). One Markdown vault, any agent.","directories":{},"_nodeVersion":"22.23.2","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.12.0_1788099174024_0.6987351689305263","host":"s3://npm-registry-packages-npm-production"}},"0.13.0":{"name":"create-ai-memory","version":"0.13.0","keywords":["ai","memory","claude","codex","gemini","cursor","opencode","cli","agent","obsidian","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.13.0","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"a500fbea05de3678648647e1199ddc9493672699","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.13.0.tgz","fileCount":19,"integrity":"sha512-LxzCecE/t5aBmC2BsRMGcEwvjo2dl7bI7gsKQT3jwEWkI6RV6cCaNnL7h36hFmHhpgJH6+tAEJEg1UxllXq7iQ==","signatures":[{"sig":"MEYCIQCTZsJyscv/OJ1HgwuTfRf9VIqD2xoU9fDD3erMyOe7bgIhAPG0jZPqX93L2JdOJXiTfudA+1XuQg/J3JPpUnY7MzKu","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"sig":"MEQCICf/jDCkirgisoqgFBtUIjvx7L823XgOShgsnGgaEg3XAiA5EiYDhoTmxLYFkZrVCj3gaQMREG9b7oRDikyFiC3Fig==","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.13.0","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":107321},"type":"commonjs","engines":{"node":">=18"},"gitHead":"7def6f89f6734c869605bc665e79899581ee9777","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"oidc:40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.0.2","description":"Persistent, agent-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode). One Markdown vault, any agent.","directories":{},"_nodeVersion":"22.23.2","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.13.0_1788146606016_0.279610800560363","host":"s3://npm-registry-packages-npm-production"}},"0.14.0":{"name":"create-ai-memory","version":"0.14.0","keywords":["agent","ai","claude","cli","codex","cursor","gemini","mcp","memory","obsidian","opencode","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.14.0","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"ai-mem-mcp":"bin/ai-mem-mcp.js","ai-mem-serve":"bin/ai-mem-serve.js","create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"226d0621ce0640696deee6f8880aa9206d2a7479","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.14.0.tgz","fileCount":22,"integrity":"sha512-rwbGRuMR6WzWvX1JTDRUcQ8zLV4C6oMU0Q7k96hwYdep93t3nCvA54rHUZNVvf0PZ3lxOnAA2hPrWVyYo3SCGw==","signatures":[{"sig":"MEYCIQDo6UA/FRIqyZol3gU0x2JfUNHaQAij2N3YLLA8CXiQIwIhAP/jcJsndlbjyfpPvm0iyAADmYmqNLEl5UOpL5WbrnRa","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"sig":"MEUCIAH577YyAp2iOJCEep6yJamWumWL4YIxLk5RD+DYoE04AiEA63EG086Zk+KfAuoLle8EA3X39VtQYGk5yf3SB2xfb5k=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.14.0","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":161116},"type":"commonjs","engines":{"node":">=18"},"gitHead":"f8a6371b80e72f90df6ef38af684cf2e85e5dd3e","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"oidc:40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.0.2","description":"Persistent, agent-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode). One Markdown vault, any agent.","directories":{},"_nodeVersion":"22.23.2","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.14.0_1788149478834_0.6607130716166323","host":"s3://npm-registry-packages-npm-production"}},"0.14.1":{"name":"create-ai-memory","version":"0.14.1","keywords":["agent","ai","claude","cli","codex","cursor","gemini","mcp","memory","obsidian","opencode","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.14.1","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"ai-mem-mcp":"bin/ai-mem-mcp.js","ai-mem-serve":"bin/ai-mem-serve.js","create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"0f020642e460d008d774e856314aa87627ec5c7d","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.14.1.tgz","fileCount":22,"integrity":"sha512-jEmlRAp8uAds5XJNS5E1jDU0uQ8+p7kocnWOqvEuzXtBtCMM+Rsj9Qw2Wla2dyYqsIpzICWOOQJPI/UFXSl1hQ==","signatures":[{"sig":"MEUCIQDR4b+f3sXvii/iWARXryYA9PHGTQiGqk3y82W4+mEoXwIgc02C/YPWc0VaGFqacxkfsYU4gjbw5TsX++8iYhVE9WQ=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"sig":"MEYCIQC9m9Q5WI4AHhQhvtmP+uFfvowdht4+eL1YTQIPnsjAzAIhAKoYNO5zNQmx4u9guDd9ZJ4GiHm3Jsm9t29nYIvVpvlB","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.14.1","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":163722},"type":"commonjs","engines":{"node":">=18"},"gitHead":"fb561709d5bfe93ea719964d987a05dd5b3d867e","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"oidc:40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.0.2","description":"Persistent, agent-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode). One Markdown vault, any agent.","directories":{},"_nodeVersion":"22.23.2","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.14.1_1788149946661_0.48718211277738854","host":"s3://npm-registry-packages-npm-production"}},"0.14.2":{"name":"create-ai-memory","version":"0.14.2","keywords":["agent","ai","claude","cli","codex","cursor","gemini","mcp","memory","obsidian","opencode","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.14.2","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"ai-mem-mcp":"bin/ai-mem-mcp.js","ai-mem-serve":"bin/ai-mem-serve.js","create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"2a38bf992b1fed18fc4b165d3fc4de13312ef264","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.14.2.tgz","fileCount":22,"integrity":"sha512-Pi00nRqSFrQAfR+8EqtZ5xvjXmeQIkbb68vnlwoNHsZB9MDZMd6ktE0N7Vscsza/ciBwu/y6Arlq6BxmgESyyQ==","signatures":[{"sig":"MEUCIQDXgtFcZi+cLeU9iOcVNg2SqfNWTb/r4toTtkMwMS+0+QIgcDN5FsC/z/oPSy9/akYZExHBEcYDfBX3jo77shhyhGw=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"sig":"MEUCIF7JXPy/e4sGN3z/g8O41UGOZqp7F4DTr6vin8aD9fGXAiEA25IeLGVu8TgrbdDStm8Cqb2AIJ41gfiIThfo+E041Qk=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.14.2","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":166515},"type":"commonjs","engines":{"node":">=18"},"gitHead":"af6329243506d84d66afb987723fa7ec61ed0222","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"oidc:40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.0.2","description":"Persistent, agent-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode). One Markdown vault, any agent.","directories":{},"_nodeVersion":"22.23.2","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.14.2_1788150437055_0.46992978690573173","host":"s3://npm-registry-packages-npm-production"}},"0.14.3":{"name":"create-ai-memory","version":"0.14.3","keywords":["agent","ai","claude","cli","codex","cursor","gemini","mcp","memory","obsidian","opencode","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.14.3","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"ai-mem-mcp":"bin/ai-mem-mcp.js","ai-mem-serve":"bin/ai-mem-serve.js","create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"468d0dabdb1ac3b11ccf07e4ada4f581e8021d8e","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.14.3.tgz","fileCount":22,"integrity":"sha512-gOlxi/V8EeKOuWjJew6yQWGpcN9lTlxlGwnO8apveH4Xd2T/zTRzkF9gnrxkvWurgiYPsxHVJJfM1zUxVEtfRw==","signatures":[{"sig":"MEYCIQDRhEcPiGh7aG4DnfWaLIpuNzT9CQQvmZVZhKYMt/VhCAIhAIPUzoc4ocBAEyX30YerRdFELDb5UZwIWkltJiw/wjS+","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"sig":"MEUCIHlnIXGRdmRpiKK0rY4+NxUhusYSPfa+Dmwr+9M3Y7ABAiEA5rHzA/KEFRwKD3BxZ/tLOGqSaoH+BhWP6KBwnk37P54=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.14.3","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":165769},"type":"commonjs","engines":{"node":">=18"},"gitHead":"31936a9419a9bdd6ff0c84f8339b6591c3967dc4","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"oidc:40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.0.2","description":"Persistent, agent-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode). One Markdown vault, any agent.","directories":{},"_nodeVersion":"22.23.2","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.14.3_1788190943693_0.8662360956151067","host":"s3://npm-registry-packages-npm-production"}},"0.15.0":{"name":"create-ai-memory","version":"0.15.0","keywords":["agent","ai","claude","cli","codex","cursor","gemini","mcp","memory","obsidian","opencode","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.15.0","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"ai-mem-mcp":"bin/ai-mem-mcp.js","ai-mem-serve":"bin/ai-mem-serve.js","create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"ced49d572514ab1b2be57ff234cc06ddbc4e00c5","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.15.0.tgz","fileCount":22,"integrity":"sha512-lXPQHlRJRxy3lMaYYeU+NcKU/IZmqritoakB6XPwl8fIFrF78UJrEPj29Jy9LEa7go3QAGLleGIU3VlAyDr4aQ==","signatures":[{"sig":"MEYCIQCvd5gxZJK6eW01enOHtPfo7nKPXJM3r5DQGFBDH27dVwIhAMd9b4ywwe3Ia5UN2VkMlMMZJtjJU7o5ojqBpPUcHxj4","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"sig":"MEUCIQChmcX3iE4RbJ+zUmwhRIMx0+JwCWNwPHNLThVClh/8+AIgL+XBIbpt/lXdQ6S70B+0Ok1G0EDuaIWu2jcxEk8adIg=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.15.0","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":208320},"type":"commonjs","engines":{"node":">=18"},"gitHead":"71894bc1e641e42321bb98c82c5f7a984b15fb57","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"oidc:40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.0.2","description":"Persistent, agent-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode). One Markdown vault, any agent.","directories":{},"_nodeVersion":"22.23.2","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.15.0_1789403650047_0.8949728951636398","host":"s3://npm-registry-packages-npm-production"}},"0.15.1":{"name":"create-ai-memory","version":"0.15.1","keywords":["agent","ai","claude","cli","codex","cursor","gemini","mcp","memory","obsidian","opencode","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.15.1","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"ai-mem-mcp":"bin/ai-mem-mcp.js","ai-mem-serve":"bin/ai-mem-serve.js","create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"d17e465c27dd14b1fb1b1f34f043a13720cd628f","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.15.1.tgz","fileCount":22,"integrity":"sha512-luKJZjPSmaK8kJccj8nhBUP83RWNTQ0WCucxhu2VRe0WPGpbyAAnDKO5uirUbg9Ca8A1Dyvk3GqS10mZxqrhsQ==","signatures":[{"sig":"MEYCIQC+xtDuIiVtiwZ3OEE4oSIJayhl66nw92nnGrfZ8SlEPgIhAKYvZeh21+GeGZAT2jWaHG2yrFcJcvKpG615/YJqoeWR","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"sig":"MEUCIQCemI3sNeoEDEOucI8SCQm5vRC01O+vVGsU3vf/zyoOqwIgave8tIiBmMeucq9fKRXzpwFa7JI7OT542zJPYEh1DJ8=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.15.1","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":209468},"type":"commonjs","engines":{"node":">=18"},"gitHead":"6f0a48bb02d7b98a52e2ce62d1d00103ae279fb9","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"oidc:40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.0.2","description":"Persistent, agent-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode). One Markdown vault, any agent.","directories":{},"_nodeVersion":"22.23.2","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.15.1_1789675479489_0.9680578547257168","host":"s3://npm-registry-packages-npm-production"}},"0.15.2":{"name":"create-ai-memory","version":"0.15.2","keywords":["agent","ai","claude","cli","codex","cursor","gemini","mcp","memory","obsidian","opencode","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.15.2","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"ai-mem-mcp":"bin/ai-mem-mcp.js","ai-mem-serve":"bin/ai-mem-serve.js","create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"75bd272463125501f46709d56dbb85c98ad272da","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.15.2.tgz","fileCount":22,"integrity":"sha512-RCiogF49XFkb4L6P/eQMu2wnIzRV2aM0UgAU6VCqCRP/yihTZPAAhhPuCJzy6UV7vSyoOPi39AY77iXL5BBxAA==","signatures":[{"sig":"MEUCIQC4rHDtcd0IeBjs5g84IorM+l/RanMyCx6NULMQARnX4gIgQizDpbAETNohs2tnK4hSxN7LZB6awvCqQzfkqpLr3gU=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"sig":"MEUCIB6wZ6rzL4EmiANHsFGDpiwMjI8W9ePbD+G3IOKK3mibAiEA8LMTxDVAIpIBkzRWpqYUD94iM0NZ/3Ghx6LCIY22NTk=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.15.2","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":210191},"type":"commonjs","engines":{"node":">=18"},"gitHead":"bce3a43fc683de238db96d989863dcc72ddb959d","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"oidc:40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.0.2","description":"Persistent, agent-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode). One Markdown vault, any agent.","directories":{},"_nodeVersion":"22.23.2","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.15.2_1789716716450_0.9849449094221636","host":"s3://npm-registry-packages-npm-production"}},"0.15.3":{"name":"create-ai-memory","version":"0.15.3","keywords":["agent","ai","claude","cli","codex","cursor","gemini","mcp","memory","obsidian","opencode","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.15.3","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"ai-mem-mcp":"bin/ai-mem-mcp.js","ai-mem-serve":"bin/ai-mem-serve.js","create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"a9f4b227a16e4397215bdbc080fb20cfa8613304","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.15.3.tgz","fileCount":22,"integrity":"sha512-M1yx2V9DKRD0FVr03OVQhpv11izdsOLB5y8CYuf9+FFCPHrHIQ6hEOu0WhPzc5O6+Gsd6l+UGwLFceY/PoEMkA==","signatures":[{"sig":"MEYCIQCztrAWoi9+fveLuwwMhpHQtJB3pbrq0M4okmIbUZ+GNgIhAKdat1ykD0eN6RW7Ov88fqIoaXNcOQOkXLI0TpuLcL/i","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"sig":"MEYCIQDVEwZK2dT0NuTUWfLl8GUnVmDAGo+plTyqdS5P26gfoAIhANODa0PzaLMn8271olznVh/5htESH4aEkk5c4dItLAJQ","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.15.3","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":211114},"type":"commonjs","engines":{"node":">=18"},"gitHead":"863666ccb1d2516aa75d97a88b5516ecf0441aee","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"oidc:40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.0.2","description":"Persistent, agent-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode). One Markdown vault, any agent.","directories":{},"_nodeVersion":"22.23.2","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.15.3_1789795053566_0.24332568431054558","host":"s3://npm-registry-packages-npm-production"}},"0.15.4":{"name":"create-ai-memory","version":"0.15.4","keywords":["agent","ai","claude","cli","codex","cursor","gemini","mcp","memory","obsidian","opencode","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.15.4","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"ai-mem-mcp":"bin/ai-mem-mcp.js","ai-mem-serve":"bin/ai-mem-serve.js","create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"aca801f942f47237b77c9fba93973709dcee1638","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.15.4.tgz","fileCount":22,"integrity":"sha512-E+dDpb5rE1LqcgCVrw/bBrbVMrKtUsaJ/K76gA4WqDlCXAaqfFouSiLiCiAYcpKki/92jLTHO/4LjihT4QzMeQ==","signatures":[{"sig":"MEYCIQDw6bdrl5jO/VtOboa57FgCuR1bGEvgnaEAZrTi3XMuyQIhAIqyBd+nII5Oz7rCI5RDNaF0X8iqTR6JQAccVPmwguma","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"sig":"MEQCIHwVVMWPYu5z0WLDwNkU1aUH2uzFAjDqNya5ej896ULYAiBYb1je9+ffnYOy1NmE3YHhj8Gir6eIVDcPP6ep82sJdQ==","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.15.4","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":211344},"type":"commonjs","engines":{"node":">=18"},"gitHead":"29f4b1517cc278f54ef018def193b35008a0319d","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"oidc:40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.0.2","description":"Persistent, agent-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode). One Markdown vault, any agent.","directories":{},"_nodeVersion":"22.23.2","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.15.4_1789795480280_0.8410902909850593","host":"s3://npm-registry-packages-npm-production"}},"0.16.0":{"name":"create-ai-memory","version":"0.16.0","keywords":["agent","ai","claude","cli","codex","cursor","gemini","mcp","memory","obsidian","opencode","omp","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.16.0","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"ai-mem-mcp":"bin/ai-mem-mcp.js","ai-mem-serve":"bin/ai-mem-serve.js","ai-mem-search":"bin/ai-mem-search","create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"7de48370a7264318ca12281cc490a5c5745a878d","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.16.0.tgz","fileCount":31,"integrity":"sha512-BBfRWMGLSxN9nzb+qXRnChKm0+2TQKjopJD9v+rIyRV8nO/eFCwj0mpyi5NpLeLYHQRFiLcF4gjpySkrDR+7bA==","signatures":[{"sig":"MEYCIQDJPAru+hW9jMrHu85sJfHPjlM60eoIbHIP3QgO6CLLzgIhALZ2S/d7wcDyqs4UA9JwnIl0Gp2bKRuPNR83+SJkglX3","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"sig":"MEQCIDevgrC4QfoH9w9QoM2FEt7zDyPposWJJbKMkqnnRUwAAiBqtSvmmwP6eWmbjxR4CA+ozqHeXgbvmZoOdEAZngOPIg==","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.16.0","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":237815},"type":"commonjs","engines":{"node":">=18"},"gitHead":"6d9fa52b5cf15deb7e64fcddc636267ae21f7547","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"oidc:40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.1.0","description":"Persistent, harness-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode, omp). One Markdown vault, any harness.","directories":{},"_nodeVersion":"22.23.2","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.16.0_1790358381051_0.8318301087423223","host":"s3://npm-registry-packages-npm-production"}},"0.16.1":{"name":"create-ai-memory","version":"0.16.1","keywords":["agent","ai","claude","cli","codex","cursor","gemini","mcp","memory","obsidian","opencode","omp","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.16.1","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"ai-mem-mcp":"bin/ai-mem-mcp.js","ai-mem-serve":"bin/ai-mem-serve.js","ai-mem-search":"bin/ai-mem-search","create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"53af413d22f3f38b02ef76f2868df5ae701cb577","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.16.1.tgz","fileCount":31,"integrity":"sha512-nBvVYncyB5PasnZSK1CPkuOPh7fYlQRCqD+MgqbKifpjIL1YubsRLdkpklq0ko2G0SHiYbzFGrBRln1aZiTisw==","signatures":[{"sig":"MEQCIAnMbHcQpu4dpKzMcCnD2at9V6ohGGSwfWufeSj4mm2+AiAhBBoOMRq37MF/9TIONeGRtLCUwtrD2YJve5aDWtRnKA==","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"sig":"MEQCIBq1m0LSUK6Lrxi1s70g3g3a3cwVM+uw79tEq8nSaQ8cAiBwHo56/cWh0lwh3UarBuqzEKebo7vQb5nvtHnTmT+tfw==","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.16.1","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":241417},"type":"commonjs","engines":{"node":">=18"},"gitHead":"745b0711e328d6d3c97f031716ac668c0e9b1b91","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"oidc:40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.1.0","description":"Persistent, harness-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode, omp). One Markdown vault, any harness.","directories":{},"_nodeVersion":"22.23.2","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.16.1_1790499879533_0.64256240220389","host":"s3://npm-registry-packages-npm-production"}},"0.16.2":{"name":"create-ai-memory","version":"0.16.2","keywords":["agent","ai","claude","cli","codex","cursor","gemini","mcp","memory","obsidian","opencode","omp","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.16.2","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"ai-mem-mcp":"bin/ai-mem-mcp.js","ai-mem-serve":"bin/ai-mem-serve.js","ai-mem-search":"bin/ai-mem-search","create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"8000d4cd83e50651826d001fd3c56da0671da72a","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.16.2.tgz","fileCount":31,"integrity":"sha512-KSiev/9oXSnb70yHOfmbtQ5tmKi0zjkZ5c4CG2tzfqC+H3GDztbcO6ZrM2oo8Th0tEVW5VEm9bYoPoZjk+dSWA==","signatures":[{"sig":"MEUCIFUavtqM+rMyrycZp++HjMQqdljmb9vpV7d9+ocHiO7rAiEA3PQTZ71Npn4UoitywG3+yYNsIcJLJf8Y93CIfsOAeis=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"sig":"MEUCIQCcANllhIufC6RkRZIV+kHZXiNUONs8mIU2ojXfy2HzoAIgQoe5yF3Wgfg6krRjtntBWXBHb/V26lZ6wNzHQZJ/MIw=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.16.2","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":242674},"type":"commonjs","engines":{"node":">=18"},"gitHead":"a233f08e2b4bdd7fca7de5f3574797617256bcd8","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.2.0","description":"Persistent, harness-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode, omp). One Markdown vault, any harness.","directories":{},"_nodeVersion":"22.23.3","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.16.2_1791124270613_0.2796910453338368","host":"s3://npm-registry-packages-npm-production"}},"0.16.3":{"name":"create-ai-memory","version":"0.16.3","keywords":["agent","ai","claude","cli","codex","cursor","gemini","mcp","memory","obsidian","opencode","omp","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.16.3","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"ai-mem-mcp":"bin/ai-mem-mcp.js","ai-mem-serve":"bin/ai-mem-serve.js","ai-mem-search":"bin/ai-mem-search","create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"93ffec3f2746f8526af09f43a98f1dd89c99f4b8","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.16.3.tgz","fileCount":31,"integrity":"sha512-7BA25dkQJTfl+wAOlgcLaExSIOEg5M/3hc5WjRn4iIRmAFIEmD0fiIndNeUlDsYkg5Yb6nwmzQ/kCRR4lryRgQ==","signatures":[{"sig":"MEYCIQCUPmkRIBjw59LR7Kxl0P/sMb7X7A4iyOuxSaSg2o7fIAIhAPPymdJpcT2lOmihkvjdsfDw9kMfD6UZNqEwPlh7K231","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"sig":"MEUCIDltb4MbBDoWomioHhGm+3VYbz7fV2Ax9YJhysqBMi5JAiEA7p02KfTzBeRtTjj0oOaQ8yNyoyPOqHGcwUkZ8/Rm4AQ=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.16.3","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":243567},"type":"commonjs","engines":{"node":">=18"},"gitHead":"1a892cfabf9d00250784e36179c6d9fb318cdd95","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.2.0","description":"Persistent, harness-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode, omp). One Markdown vault, any harness.","directories":{},"_nodeVersion":"22.23.3","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.16.3_1791655283586_0.7890352712912061","host":"s3://npm-registry-packages-npm-production"}},"0.16.4":{"name":"create-ai-memory","version":"0.16.4","keywords":["agent","ai","claude","cli","codex","cursor","gemini","mcp","memory","obsidian","opencode","omp","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.16.4","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"ai-mem-mcp":"bin/ai-mem-mcp.js","ai-mem-serve":"bin/ai-mem-serve.js","ai-mem-search":"bin/ai-mem-search","create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"90ded17be65f37e4792e960eb5430d4c0a3cc0d9","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.16.4.tgz","fileCount":31,"integrity":"sha512-4vTA3DNBxJt3p9w4PleJJnNoI735UQ/krwF3+/fclJtvCmuagzw0UCiqGt6jjjIBvnGLfWB1TMN3mQvRbJXLyA==","signatures":[{"sig":"MEYCIQCN8zgJQAqwe5SXj/VuKEEp+8/FvAJhzZFYrEZmkM0OOQIhAJlBsCasc7Gf0mMN54nvcgEo9vtHmbGmbzterVBZyRmH","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"sig":"MEQCIFVkheNTGXT4HOMi9Yya6D246RQZhB7r88t3Q9UN4O2MAiByOrEWUc1SLIs4MJqLE1atUmwKTkkgSv6WL/O01DhM3A==","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.16.4","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":246674},"type":"commonjs","engines":{"node":">=18"},"gitHead":"a759b5a6cf0814b5b021c6efeb71fecb08e0af43","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.2.1","description":"Persistent, harness-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode, omp). One Markdown vault, any harness.","directories":{},"_nodeVersion":"22.23.4","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.16.4_1793695119781_0.2812813238465025","host":"s3://npm-registry-packages-npm-production"}},"0.16.5":{"name":"create-ai-memory","version":"0.16.5","keywords":["agent","ai","claude","cli","codex","cursor","gemini","mcp","memory","obsidian","opencode","omp","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.16.5","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"ai-mem-mcp":"bin/ai-mem-mcp.js","ai-mem-serve":"bin/ai-mem-serve.js","ai-mem-search":"bin/ai-mem-search","create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"912c4dc6507397a850de58cdf476881e3a648173","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.16.5.tgz","fileCount":31,"integrity":"sha512-YKBaf+xLpVZQSCRZYLlC64uTFJquNESkqk6sy2P13ZrPOjtOuNsWUe3RkxCRSERA5CqqvKN14ac2saRTDH0ezw==","signatures":[{"sig":"MEUCIAGnjHuBoX85YTKcMOAbNPeIDGhqXshmXRu2/U+3xj/4AiEA/WECdCw5eYpwFiFXmAqKU/lobngW3gaW0t1V0HyX+sA=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"sig":"MEQCICnj/0KS/jm4AvkUZRJsJoSp2Gt8DojtHEDUVi0hP24AAiBV3/1TsW6YpneBWkBKq+SMnTTJu0se/6Yz/LVbP2tzzg==","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.16.5","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":246959},"type":"commonjs","engines":{"node":">=18"},"gitHead":"e9d4464f2a8f2c387f975af3aae07fa74e261465","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.2.1","description":"Persistent, harness-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode, omp). One Markdown vault, any harness.","directories":{},"_nodeVersion":"22.23.4","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.16.5_1793711239615_0.5279645196925603","host":"s3://npm-registry-packages-npm-production"}},"0.17.0":{"name":"create-ai-memory","version":"0.17.0","keywords":["agent","ai","claude","cli","codex","cursor","gemini","mcp","memory","obsidian","opencode","omp","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.17.0","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"ai-mem-mcp":"bin/ai-mem-mcp.js","ai-mem-serve":"bin/ai-mem-serve.js","ai-mem-search":"bin/ai-mem-search","create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"d246e232f4633dfb4481f2590296711521d5a5c6","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.17.0.tgz","fileCount":31,"integrity":"sha512-qwQtk6hTCZ7KWPX1ZmQzjU6PXNiEiTedoNPMLKJnfXBmcfHkwrdFwFjma8qreVgHzwSg/z4shszCBdT+zeeIlg==","signatures":[{"sig":"MEQCIAvactFOPLl+FixhfJmYd3ttgA2zIaNnk95EvWlb9YVpAiAVQ1myVeUH72KGrP50g/UMiODRkfrVpSbby/vqVvDojA==","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"sig":"MEUCIB5wpwtKQkNy8CrpMg+tveD2wqGPuqH5Y45eYOyaninQAiEAs+sL6P85Rvg6s+LPVP9HtKVGqVmoj1OH+gUdoml7eQo=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.17.0","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":250354},"type":"commonjs","engines":{"node":">=18"},"gitHead":"05263696a83040bbdb14a3f17f0420e77d0fb734","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.2.1","description":"Persistent, harness-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode, omp). One Markdown vault, any harness.","directories":{},"_nodeVersion":"22.23.4","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.17.0_1793720629060_0.6425214010124131","host":"s3://npm-registry-packages-npm-production"}},"0.18.0":{"name":"create-ai-memory","version":"0.18.0","keywords":["agent","ai","claude","cli","codex","cursor","gemini","mcp","memory","obsidian","opencode","omp","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.18.0","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"ai-mem-mcp":"bin/ai-mem-mcp.js","ai-mem-serve":"bin/ai-mem-serve.js","ai-mem-search":"bin/ai-mem-search","create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"eadcd47a86527a63116610b216277e6b5bef2c60","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.18.0.tgz","fileCount":31,"integrity":"sha512-G4F6J0hn+rbe8qmAYigPvXNiffG6BWIZMaZo2X/b8pYyIexhr2i+vKw30wplUJeRJO2/o4RRaJiM+YqDd4GBwQ==","signatures":[{"sig":"MEUCIAHCMlh/4dz8C0USGOvpBWtQEWovXnnvu7y1+I0HOgIzAiEAuSr6iCwBLD7+Qa/Fw1FFUmfluGcYgWYytco1Rvp3ioo=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"sig":"MEUCIHalfuaM83cPXXKGSLgy/Se/K0ySIoxdJ0mZatl9W3SlAiEAgDATu9pqwyKPbGWUQ6Q5a8EfxhsADP2WgHPrwPitDKw=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.18.0","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":255918},"type":"commonjs","engines":{"node":">=18"},"gitHead":"d589d9323d95f6f1c62731d0add81d199959e334","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.2.2","description":"Persistent, harness-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode, omp). One Markdown vault, any harness.","directories":{},"_nodeVersion":"22.23.4","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.18.0_1794650610533_0.3903358845748994","host":"s3://npm-registry-packages-npm-production"}},"0.18.1":{"name":"create-ai-memory","version":"0.18.1","keywords":["agent","ai","claude","cli","codex","cursor","gemini","mcp","memory","obsidian","opencode","omp","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.18.1","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"ai-mem-mcp":"bin/ai-mem-mcp.js","ai-mem-serve":"bin/ai-mem-serve.js","ai-mem-search":"bin/ai-mem-search","create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"7a11aedbe8c90a2c4ccdc75ee19956fe688f62cf","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.18.1.tgz","fileCount":31,"integrity":"sha512-Rss4Xeq/aSC8NVyXB6coHiXWFvDSYIa1k0DnHTdXpYC4SXaecONETjQ4pXSw9WLnL7wqg3wUKUgOHjy3J4Db6w==","signatures":[{"sig":"MEUCIDtlhc7rv7w2ovykUHxZNk7jg037b6ehMglCTgADmKiIAiEAirp0FLn890DZMF0uD4lPJPrhXhWOEQVGEHwMUTPh2hE=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"sig":"MEUCIC5+5JsD9dTtmHTNEM8mc+6L7sJDODIpB/tXy1rg5C6LAiEAxMns7L+eDAnqFeoH894FuOCKiDhLPzpfRl969Iy763I=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.18.1","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":257886},"type":"commonjs","engines":{"node":">=18"},"gitHead":"e1ba1f8c619a0424a266c7c828f15ecf2cd0549a","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.2.2","description":"Persistent, harness-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode, omp). One Markdown vault, any harness.","directories":{},"_nodeVersion":"22.23.4","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.18.1_1794652067372_0.05120290741669753","host":"s3://npm-registry-packages-npm-production"}},"0.18.2":{"name":"create-ai-memory","version":"0.18.2","keywords":["agent","ai","claude","cli","codex","cursor","gemini","mcp","memory","obsidian","opencode","omp","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.18.2","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"ai-mem-mcp":"bin/ai-mem-mcp.js","ai-mem-serve":"bin/ai-mem-serve.js","ai-mem-search":"bin/ai-mem-search","create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"e4678ef37665d4f7ac30665998987702b7795250","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.18.2.tgz","fileCount":31,"integrity":"sha512-cD4biDKe9i1wGyO6LxFUdHyNAIfSOzmOg8aXtSJ1y9DVOQ+Bt1Ox2Xm6cB/G5Ux9MG8PzuoQERyQekFURWZGuQ==","signatures":[{"sig":"MEQCIAf7XVmBjWO/4+I6JIB3HrjBGSYqSazMw1Ug9Jo7OeKDAiAjjlEvXB47wIuetNOu2WM1TgiSiPnbf99VvNbbMKLwFw==","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"sig":"MEUCICbZV9d3P/b4IMD3YDFMsP+H0w+/Z2c7F7q27ovn8f5dAiEA00OGBHLCMwfXuHXXFHt1bJtZC4BP/AxEN81ByQ0VFq4=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.18.2","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":259695},"type":"commonjs","engines":{"node":">=18"},"gitHead":"9cdb33712ab577d0ffc024ea1d67e48662688016","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.2.2","description":"Persistent, harness-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode, omp). One Markdown vault, any harness.","directories":{},"_nodeVersion":"22.23.4","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.18.2_1794654646616_0.2511381595350354","host":"s3://npm-registry-packages-npm-production"}},"0.19.0":{"name":"create-ai-memory","version":"0.19.0","keywords":["agent","ai","claude","cli","codex","cursor","gemini","mcp","memory","obsidian","opencode","omp","zsh"],"author":{"name":"Ram Christopher Baarde"},"license":"MIT","_id":"create-ai-memory@0.19.0","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"homepage":"https://github.com/rambaarde/create-ai-memory#readme","bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"bin":{"ai-mem-mcp":"bin/ai-mem-mcp.js","ai-mem-serve":"bin/ai-mem-serve.js","ai-mem-search":"bin/ai-mem-search","create-ai-memory":"bin/create-ai-memory.js"},"dist":{"shasum":"ab4aa09c2b5b21e5724166fe63e8adb4555a7994","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.19.0.tgz","fileCount":31,"integrity":"sha512-SstCcrr9eO93wnkfnHA5vZkClLNzqnTF2Fz2mWV7m5niXAU4PCXWI5uoyXMuTZcGtyubQfpFQMMYoN3kBYTJeA==","signatures":[{"sig":"MEUCIQDMyVt8uMDvVEWMi7hsVGl0aLgMv74Md4ntS1BlMmelTgIgEbx62n0qPdPfVkSAq+i8bfFfpBaBX27ZDXKdbwB6QUo=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"sig":"MEUCIDOo48HdafapCZoexi6r0y7f5MDl3ZT1Z7IWBU3CpoiuAiEA9yAhSxnubJQBJwgw4iNgYRnkzLjZlRU7N+LPKBhAe7E=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.19.0","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":265226},"type":"commonjs","engines":{"node":">=18"},"gitHead":"e46d1e747344ee0a27daf8a15671bde00de028df","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.2.2","description":"Persistent, harness-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode, omp). One Markdown vault, any harness.","directories":{},"_nodeVersion":"22.23.4","_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/create-ai-memory_0.19.0_1794856451019_0.27695451774182844","host":"s3://npm-registry-packages-npm-production"}},"0.19.1":{"_id":"create-ai-memory@0.19.1","bin":{"ai-mem-mcp":"bin/ai-mem-mcp.js","ai-mem-serve":"bin/ai-mem-serve.js","ai-mem-search":"bin/ai-mem-search","create-ai-memory":"bin/create-ai-memory.js"},"bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"dist":{"shasum":"a48b75a8e3c9c6ffe388a22a1c803a7e99db246b","tarball":"https://registry.npmjs.org/create-ai-memory/-/create-ai-memory-0.19.1.tgz","fileCount":34,"integrity":"sha512-8P+jUDI8lMlmBcyqyxoXco1yTrR+uLT5IzWk7RPLQIZ4JQcx5jG3zSyzb1UfEftnDZ4Kk3+H+Gjx+d/yd5aqnQ==","signatures":[{"sig":"MEQCIEEHNzzVZPqkXsV6fj2AF7nJMD2dtwIBogztc8RNC0JzAiBPSqLBrOkuAGMWDGkG+22vmzLuc05oeuMe6d/oUvfNBQ==","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U","sig":"MEYCIQDsFyXyZiBaqDTrN2QDIt4W8dFVaApzgtyTsiy6cR4riAIhALP4qNLjj0Q/SUJWozUg9WFf1cFunXon2joQa5n/atCW"}],"attestations":{"url":"https://registry.npmjs.org/-/npm/v1/attestations/create-ai-memory@0.19.1","provenance":{"predicateType":"https://slsa.dev/provenance/v1"}},"unpackedSize":266449},"name":"create-ai-memory","type":"commonjs","author":{"name":"Ram Christopher Baarde"},"engines":{"node":">=18"},"gitHead":"f317923feb99a51e12e835cbd41f982abaf97b1a","license":"MIT","scripts":{"test":"zsh tests/run.sh","prepublishOnly":"npm test"},"version":"0.19.1","_npmUser":{"name":"GitHub Actions","email":"npm-oidc-no-reply@github.com","trustedPublisher":{"id":"github","oidcConfigId":"40b68da9-7f05-46eb-a4d8-9756009d4a52"}},"homepage":"https://github.com/rambaarde/create-ai-memory#readme","keywords":["agent","ai","claude","cli","codex","cursor","gemini","mcp","memory","obsidian","opencode","omp","zsh"],"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"_npmVersion":"12.2.2","description":"Persistent, harness-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode, omp). One Markdown vault, any harness.","directories":{},"maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"_nodeVersion":"22.23.4","_hasShrinkwrap":false,"_npmOperationalInternal":{"host":"s3://npm-registry-packages-npm-production","tmp":"tmp/create-ai-memory_0.19.1_1794901168136_0.8959824706164408"}}},"time":{"created":"2026-07-13T17:23:46.563Z","modified":"2026-11-17T07:39:28.538Z","0.1.0":"2026-07-13T17:23:46.822Z","0.1.1":"2026-07-13T18:26:07.313Z","0.1.2":"2026-07-13T18:36:23.014Z","0.2.0":"2026-08-27T16:58:16.152Z","0.2.1":"2026-08-28T03:31:09.778Z","0.3.0":"2026-08-28T07:09:15.905Z","0.4.0":"2026-08-28T07:24:46.247Z","0.5.0":"2026-08-28T08:32:10.169Z","0.6.0":"2026-08-28T21:31:40.737Z","0.7.1":"2026-08-29T01:19:08.469Z","0.8.0":"2026-08-30T05:02:46.079Z","0.8.1":"2026-08-30T07:11:50.330Z","0.8.2":"2026-08-30T07:20:35.516Z","0.8.3":"2026-08-30T07:32:23.415Z","0.9.0":"2026-08-30T07:45:07.874Z","0.9.1":"2026-08-30T07:54:22.130Z","0.10.0":"2026-08-30T13:14:49.887Z","0.11.0":"2026-08-30T13:25:59.448Z","0.11.1":"2026-08-30T13:34:40.496Z","0.11.2":"2026-08-30T13:54:35.213Z","0.11.3":"2026-08-30T14:03:02.553Z","0.12.0":"2026-08-30T14:12:54.112Z","0.13.0":"2026-08-31T03:23:26.100Z","0.14.0":"2026-08-31T04:11:18.910Z","0.14.1":"2026-08-31T04:19:06.749Z","0.14.2":"2026-08-31T04:27:17.145Z","0.14.3":"2026-08-31T15:42:23.781Z","0.15.0":"2026-09-14T16:34:10.145Z","0.15.1":"2026-09-17T20:04:39.599Z","0.15.2":"2026-09-18T07:31:56.567Z","0.15.3":"2026-09-19T05:17:33.666Z","0.15.4":"2026-09-19T05:24:40.395Z","0.16.0":"2026-09-25T17:46:21.142Z","0.16.1":"2026-09-27T09:04:39.622Z","0.16.2":"2026-10-04T14:31:10.709Z","0.16.3":"2026-10-10T18:01:23.693Z","0.16.4":"2026-11-03T08:38:39.889Z","0.16.5":"2026-11-03T13:07:19.708Z","0.17.0":"2026-11-03T15:43:49.158Z","0.18.0":"2026-11-14T10:03:30.655Z","0.18.1":"2026-11-14T10:27:47.474Z","0.18.2":"2026-11-14T11:10:46.686Z","0.19.0":"2026-11-16T19:14:11.220Z","0.19.1":"2026-11-17T07:39:28.235Z"},"bugs":{"url":"https://github.com/rambaarde/create-ai-memory/issues"},"author":{"name":"Ram Christopher Baarde"},"license":"MIT","homepage":"https://github.com/rambaarde/create-ai-memory#readme","keywords":["agent","ai","claude","cli","codex","cursor","gemini","mcp","memory","obsidian","opencode","omp","zsh"],"repository":{"url":"git+https://github.com/rambaarde/create-ai-memory.git","type":"git"},"description":"Persistent, harness-agnostic session memory for AI coding CLIs (Claude Code, Codex, Gemini, Cursor, opencode, omp). One Markdown vault, any harness.","maintainers":[{"name":"rambaarde","email":"ramchrist20@gmail.com"}],"readme":"<div align=\"center\">\n\n# create-ai-memory\n\n**The persistent memory layer for AI coding agents. One Markdown vault, every CLI.**\n\nYour agent forgets everything the moment a session ends. create-ai-memory gives Claude\nCode, Codex, Gemini, Cursor, and opencode a shared second brain: a plain-Markdown\nvault that carries your profile, each project's context, and where you left off\ninto every new session, on its own.\n\nNo daemon. No database. No API key. Just zsh and Markdown you can read.\n\n```sh\nnpm create ai-memory@latest\n```\n\n[![npm](https://img.shields.io/npm/v/create-ai-memory?color=cb3837&label=npm)](https://www.npmjs.com/package/create-ai-memory)\n[![ci](https://img.shields.io/github/actions/workflow/status/rambaarde/create-ai-memory/ci.yml?label=ci)](https://github.com/rambaarde/create-ai-memory/actions/workflows/ci.yml)\n[![release](https://img.shields.io/github/actions/workflow/status/rambaarde/create-ai-memory/publish.yml?label=release)](https://github.com/rambaarde/create-ai-memory/actions/workflows/publish.yml)\n![zsh](https://img.shields.io/badge/shell-zsh-89e051)\n![deps](https://img.shields.io/badge/runtime%20deps-0-blue)\n![license](https://img.shields.io/badge/license-MIT-blue)\n![PRs](https://img.shields.io/badge/PRs-welcome-orange)\n\n<br>\n\n<img src=\"https://raw.githubusercontent.com/rambaarde/create-ai-memory/main/assets/demo.gif\" alt=\"Installing create-ai-memory with npm create ai-memory@latest\" width=\"820\">\n\n</div>\n\n---\n\n## The problem\n\nEvery AI coding session starts from zero. The moment a thread ends, the agent\nforgets what was built, why each choice was made, and where you left off.\n\nYou pay it again on the next run: re-explaining the stack, re-litigating\nsettled decisions, losing \"where I was\". Switching agents makes it worse —\neach CLI is its own island, so what Claude learned never reaches Codex.\n\n## How it works\n\nKeep the memory outside the chat, in plain Markdown on disk, and inject it\ninto whichever agent you launch. A thread is disposable; the vault is\npermanent. Because it is just files, the same vault opens in\n[Obsidian](https://obsidian.md) with graph view and backlinks — though\nnothing here requires it.\n\nThree layers, each injected at the right scope:\n\n| Layer | Lives in | Injected | Holds |\n|---|---|---|---|\n| **Global** | `_Global_Profile.md`, `_Standards.md` | every session | your rules, commit policy |\n| **About me** | `_about_me/*.md` | every session: hard \"no\"s in full, the rest as an index | non-negotiables, wants, interests, values, expertise |\n| **Project** | `_projects/<repo>.md` | sessions in that repo | purpose, architecture, constraints, decisions |\n| **Session** | `_session_logs/<repo>/<timestamp>.md` | next session as carryover | what changed, blockers, next steps |\n\nThe prior session's outcome is **inlined into the opening prompt**, not left\nas a path the agent is told to go read. One environment variable,\n`AI_MEM_ROOT`, points at the vault, so the whole system moves between\nmachines by pointing at the same folder.\n\n## The full loop\n\nAgents read the vault through `ai-mem.zsh` at launch and write to it through a Stop\nhook at close. You read the same files in Obsidian, or through `ai-mem-search` and\n`ai-mem-lint`. Overnight, `ai-mem-sleep` tidies the same folder. Neither side gets a\nprivileged interface — it's the same folder either way.\n\n```mermaid\nsequenceDiagram\n    autonumber\n    actor You\n    participant Shell as claude-start\n    participant Agent as Claude Code\n    participant Vault as _Ai_Memory\n    participant Sleep as ai-mem-sleep\n    participant Backup as git backup (optional)\n\n    You->>Shell: claude-start\n    Shell->>Vault: read _Global_Profile.md + _Standards.md\n    Vault-->>Shell: full text\n    Shell->>Vault: read prior session's Session Outcome\n    Vault-->>Shell: Summary / Decisions / Blockers / Next\n    Shell->>Vault: create session log (project + previous wikilinks)\n    Shell->>Agent: launch, all of it as the first message\n    Note over You,Agent: the session's work happens here\n    You->>Agent: ai-note \"...\" (optional, mid-session)\n    Agent->>Vault: append to Live Notes\n    Note over Agent: session ends -- Stop hook fires\n    Agent->>Vault: rewrite Auto Session Log (branch, commits, changes)\n    Agent->>Backup: commit + push, if the vault is git-backed\n    Note over Vault: nothing overwritten -- a fresh file every session\n    Note over Sleep: overnight, unattended -- installed by ai-mem-sleep-schedule\n    Sleep->>Vault: archive unfilled logs after a day, logs older than 90d (newest 5 filled kept)\n    Sleep->>Vault: lint links, rank links to unwritten notes, flag consolidation candidates\n    Sleep->>Backup: commit + push the archive move, if the vault is git-backed\n    Note over Vault,Sleep: _lessons/ are never archived and never pruned\n    You->>Shell: claude-start (next time)\n    Shell->>Vault: read prior session again\n    Note over Shell,Vault: now includes what just happened\n```\n\nEverything through step 10 touches only files on your machine. Step 11 and\nstep 14 are the only network calls in the whole loop, and both only exist if\nyou set up a git-backed vault yourself; skip that and everything stays local,\nalways. The overnight pass (steps 12-14) is the pruning half of the loop --\nwhat decays, what is consolidated, and what is never touched is set out in\n[Memory hygiene](#memory-hygiene).\n\n## See a full session\n\n```console\n$ cd ~/code/checkout-api\n$ claude-start\n  Use terse output this session? [y/N] n\n\n  # Claude launches pre-loaded with:\n  #   • your profile + coding standards + commit policy   (global)\n  #   • checkout-api: purpose, architecture, decisions    (project)\n  #   • \"Next: wire the refund webhook\"                   (last session's carryover)\n\n… you build the refund webhook, make a few commits …\n\n$ ai-note \"refund webhook live; still need idempotency keys\"   # jot mid-session\n\n# On exit, a hook stamps the session log with the branch, the commits you made,\n# and anything uncommitted. Tomorrow's claude-start picks up exactly there.\n```\n\nNo copy-pasting context. No re-explaining the stack. No \"where were we.\"\n\n## Quickstart\n\n```sh\nnpm create ai-memory@latest     # copies the tool in and runs the setup, no git clone\nexec zsh\n```\n\nThen, from inside any git repo:\n\n```sh\nclaude-start                    # or codex-start / agy-start / cursor-start / opencode-start\n```\n\nThe agent opens already knowing your standards, this project, and where you left\noff last time.\n\n## Table of contents\n\n<table>\n<tr>\n<td valign=\"top\" width=\"33%\">\n\n**Overview**\n\n- [The problem](#the-problem)\n- [How it works](#how-it-works)\n- [See a full session](#see-a-full-session)\n- [What you get](#what-you-get)\n\n**Getting started**\n\n- [Quickstart](#quickstart)\n- [Benchmarks](#benchmarks)\n- [Install](#install)\n- [Commands](#commands)\n- [How search works](#how-search-works)\n- [What it costs](#what-it-costs-and-what-it-cannot-do)\n- [Memory hygiene](#memory-hygiene)\n\n</td>\n<td valign=\"top\" width=\"33%\">\n\n**Reference**\n\n- [Session skills](#session-skills-optional)\n- [Your vault](#your-vault)\n- [Open Knowledge Format](#open-knowledge-format)\n- [Add another agent](#add-another-agent)\n- [Integrations](#integrations)\n- [Graph view](#graph-view)\n- [GUI clients (MCP)](#gui-clients-mcp)\n- [Configuration](#configuration)\n\n</td>\n<td valign=\"top\" width=\"33%\">\n\n**Project**\n\n- [Why not just CLAUDE.md?](#why-not-just-claudemd)\n- [Troubleshooting](#troubleshooting)\n- [Why plain files](#why-plain-files)\n- [Tests](#tests)\n- [FAQ](#faq)\n- [Roadmap](#roadmap)\n- [Contributing](#contributing)\n- [License](#license)\n\n</td>\n</tr>\n</table>\n\n## What you get\n\n| | |\n|---|---|\n| **Cross-agent memory** | One vault serves Claude Code, Codex, Gemini, Cursor, and opencode — and GUI clients like Claude Desktop through an MCP server. Context earned in one reaches the next. |\n| **Automatic carryover** | A Stop hook writes the branch, the commits you made, and uncommitted changes into the session log, so tomorrow's run resumes where today's ended. |\n| **Per-project context** | Each git repo gets its own note for purpose, architecture, and decisions, injected only for that project. |\n| **Your rules, everywhere** | A global profile and standards note ride along in every session, on every project. |\n| **Session skills you define** | Register your own y/n launch options (terse output, design review, minimal-code). create-ai-memory ships none; they are yours. |\n| **Open agent model** | Adapters are three lines. Add opencode, aider, or anything with a CLI without touching the core. |\n| **Obsidian-native** | The vault is plain Markdown, so it opens as an Obsidian second brain with graph view and backlinks, or as plain files with grep. |\n| **Guardrails built in** | Every write is checked to stay inside the vault, and a commit hook refuses commits made without the vault context loaded. |\n| **Zero runtime deps** | No daemon, no database, no API key, no server. It runs in your shell. |\n\n## Benchmarks\n\nMeasured on a real 495-note vault (4.9 MB of Markdown), macOS, warm cache.\nEvery number below is reproducible with the command beside it.\n\n**Tokens** — what the agent pays, every session\n\n| | before | after | |\n|---|---|---|---|\n| Launch prompt | 7,207 | **4,046** | −44%, mirrored notes injected once ([why](#mirrored-notes-are-not-injected-twice)) and inlined notes capped ([why](#every-path-into-the-prompt-is-bounded)) |\n| Search result, common term | 10,234 | **1,752** | −83%, bounded output ([why](#how-search-works)) |\n| Lesson index, 96 lessons | — | **+829** | ~8 tokens per lesson, names only ([why](#lessons-are-indexed-not-injected)) |\n\n```sh\n# launch prompt\nai-context | wc -c            # chars; divide by ~4 for tokens\n# search output, bounded vs not\nai-mem-search postgres | wc -c\nAI_MEM_SEARCH_LIMIT=999999 ai-mem-search postgres | wc -c\n```\n\n**Speed**\n\n| | |\n|---|---|\n| Search, common term, 501-note vault | **0.07 s** |\n| Search at 2,000 / 10,000 notes | 1.3 s / 7.2 s — linear, see [limits](#what-it-costs-and-what-it-cannot-do) |\n| Shell startup cost of the module | **< 0.01 s** |\n| Recency sort, v0.11.0 regression fix | 25.4 s → **0.54 s** (47×) |\n\n```sh\ntime ai-mem-search postgres >/dev/null\ntime zsh -c 'source shell/ai-mem.zsh'\n```\n\nThe 47× is worth stating plainly because it was self-inflicted. The sort\nspawned two subprocesses per matched line -- about 18,000 processes for one\nquery -- while the `grep` underneath it took 0.16 s. Benchmarking `grep`\nalone said the command was fast; benchmarking the command said otherwise.\nA regression guard now bounds it, with the bound set by measuring both\nimplementations rather than guessing.\n\n**Footprint**\n\n| | |\n|---|---|\n| Runtime dependencies | **0** |\n| Package | **35 kB** (99.7 kB unpacked, 19 files) |\n| Index, daemon, database, embeddings | **none** — a note is searchable the moment it is written |\n| Tests | **305**, no network, no framework |\n\n```sh\nnpm pack --dry-run && zsh tests/run.sh\n```\n\n## Install\n\nPick whichever fits how you manage your shell. All paths end at the same place.\n\n**npm** (no git clone; the tool is bundled in the package):\n\n```sh\nnpm create ai-memory@latest         # into ~/ai-memory, then runs the setup\nnpx create-ai-memory ~/code/ai-memory   # or a directory you choose\n```\n\nPackage: [npmjs.com/package/create-ai-memory](https://www.npmjs.com/package/create-ai-memory)\n\n**zsh plugin manager:**\n\n```zsh\n# zinit\nzinit light rambaarde/create-ai-memory\n\n# antidote (in your plugins file)\nrambaarde/create-ai-memory\n\n# oh-my-zsh: clone into custom/plugins, then add create-ai-memory to plugins=(...)\n```\n\nPlugin-manager installs only source the module. That is fine: the vault\nauto-scaffolds from the shipped templates on first use, so `install.sh` is\noptional. Set `AI_MEM_ROOT` in `~/.zshrc` first if you do not want the default\n`~/.ai-memory/_Ai_Memory`.\n\n> **zsh only.** The module uses `print -r`, `${(s:|:)}`, and associative arrays. A bash\n> port is welcome as a PR; see [Roadmap](#roadmap).\n\n## Commands\n\n| Command | What it does |\n|---|---|\n| `claude-start` · `codex-start` · `agy-start` · `gemini-start` · `cursor-start` · `opencode-start` · `omp-start` | Launch a harness with full vault context and the session-skill picker |\n| `ai-start [project]` | Prepare the session (project note and fresh log) without launching an agent |\n| `ai-context [project]` | Print the vault context block for the current repo, and arm the git commit guard |\n| `ai-note <text>` | Append a timestamped note to today's session log while you work |\n| `ai-lesson <topic-slug> <problem> <solution>` | Append a dated Problem/Solution entry to a cross-project `_lessons/<topic-slug>.md` -- decisions, mistakes, solutions worth recalling outside the current project |\n| `ai-about-me [--review] [harness]` | Fill `_about_me/` by interview. The agent asks one question at a time, asks about real past events before hypotheticals, asks \"why does that matter?\" to find the reason behind each rule, and writes a note only after you approve its draft. `--review` proposes rules found in 2 or more of your lessons and session logs; you accept, change or reject each one. The harness defaults to the first in `AI_MEM_HARNESSES`. When `_about_me/` holds only placeholders, the next launch tells you once, on stderr, never the agent |\n| `ai-mem-ingest --source-id <id> --title <t> --date <ISO> [...]` | File a distilled external artifact -- a meeting, a recorded call, a video, a voice note -- as one note in `_transcripts/`. Writes the summary, decisions, action items, attendee `[[wikilinks]]` and a link to the source; it never embeds the raw transcript body, because a full transcript is bulk the agent pays for and rarely needs. Re-ingesting the same `--source-id` updates that one note instead of filing a second copy. `ai-mem-search` covers `_transcripts/` with no extra path |\n| `ai-mem-lint [--fix]` | Check the vault's links: orphaned session logs, dangling `previous` links (a link to a log in the project's `_archive/` resolves, as it does in Obsidian), unreferenced project notes, and notes missing the `type:` field. Links to notes not written yet (`[[Name]]` with no `Name.md`) are **not** errors -- Obsidian creates the note when the link is clicked -- so they get one ranked line instead: `most linked: [[checkout-api]] 17, ...`. A target many notes link to is a missing note worth writing. `[[...]]` in code and `{{placeholders}}` are ignored. `--fix` backfills `type:` into session logs written before the field existed |\n| `ai-mem-recall` | Does the memory get used? Every whole-vault `ai-mem-search` appends the lessons it showed (time, project, slug, query) to `.ai-mem/recall.tsv` in the vault. `ai-mem-recall` reads that log and the dated entries in `_lessons/`: how many lessons searches showed, and for each **repeat** entry (the same problem came back) whether a search showed that lesson in the 24 hours before. A repeat no search showed is a **recall miss**: the answer was in the vault and the agent never saw it. The log holds your search queries and is backed up with the vault |\n| `ai-mem-search <term> [project]` | Case-insensitive literal search across the vault (or one project's logs). Lessons rank first, and within them by *reinforcement* -- a lesson recalled more often (more dated entries) ranks above a once-seen one -- then by date; archived logs (see `ai-mem-sleep`) are skipped. Paths print relative to a root stated once in the header. Output is capped (`AI_MEM_SEARCH_LIMIT`, default 25) with an explicit `N hidden` notice, because the usual caller is an agent with a finite context window. Also resolves any `[[wikilink]]` on a matched line to its project note -- one hop out along the graph, always on, not a flag to remember |\n| `ai-mem-sleep [--apply]` | The vault's \"bedtime\" pass: archive stale session logs, flag consolidation candidates, and lint, in one run. Session logs older than `AI_MEM_SLEEP_DAYS` (default 90) move out of the hot search path -- always keeping the newest `AI_MEM_SLEEP_KEEP` (default 5) filled logs per project -- while `_lessons/` are never touched, because durable failure->fix knowledge is exactly what must survive. An **unfilled** log -- the template with nothing written, plus at most an Auto Session Log that saw no commits and a clean tree -- is archived once it is dated before yesterday, and never takes a KEEP slot: most sessions close without a note, and counting them pushed every real log out of the hot set. Dry-run by default; `--apply` moves logs into each project's `_archive/` (reversible via git) and backs up |\n| `ai-mem-sleep-schedule [--install\\|--uninstall] [--at HH:MM]` | Run the bedtime pass on its own, nightly (default 03:00). macOS installs a launchd agent; other platforms a crontab entry. The scheduled job sets `AI_MEM_ROOT` and sources the module first, since a cron/launchd shell does not read `~/.zshrc`. On macOS, a vault in `~/Documents`, `~/Desktop`, `~/Downloads` or iCloud Drive gets a warning: macOS can deny a launchd job read access there, and the pass then does nothing. Check the job log after its first run. Dry-run by default -- prints exactly what it would install; `--install` schedules it, `--uninstall` removes it |\n| `ai-mem-serve [port] [--no-open]` | Open the vault as a browsable graph on `127.0.0.1`. Agents run this for you when you ask to see your memory (see [Graph view](#graph-view)) |\n| `ai-mem-vault-backup` | Commit and push the vault if it's git-backed. `ai-note`/`ai-lesson` already call this; use it directly after editing a session log or project note by hand |\n\n`codex-note` and `codex-lesson` are aliases of `ai-note` and `ai-lesson`, for\nmuscle memory inside a Codex session; they behave identically.\n\nProject is auto-resolved from the current git repo; pass a name to override.\n\n## How search works\n\nReads your notes directly, every time. No index, no daemon, no embeddings —\nnothing to rebuild and nothing that can go stale. A note is findable the\nsecond you write it.\n\n```mermaid\nflowchart TD\n    A([\"You look for a word\"]) --> B{\"Did you name<br/>a project?\"}\n    B -->|\"no\"| C[\"Look through every note<br/>outside _archive/\"]\n    B -->|\"yes\"| D[\"Look only inside<br/>that project\"]\n    C --> E[\"Read them all.<br/>Capitals don't matter.<br/>Punctuation means itself.\"]\n    D --> E\n    E -->|\"found nothing\"| F[\"<b>Says so, plainly.</b><br/>Never a blank screen you<br/>could mistake for 'none exist'\"]\n    E -->|\"found something\"| G[\"Newest first\"]\n    G --> H[\"Show a few from each note<br/>and say how many more there are\"]\n    H --> I[\"Follow any linked note one step,<br/>with a one-line summary of it\"]\n    I --> J([\"What you read\"])\n\n    classDef ask fill:#0d9488,stroke:#0f766e,color:#fff\n    classDef care fill:#b45309,stroke:#92400e,color:#fff\n    classDef done fill:#1e3a8a,stroke:#1e40af,color:#fff\n    class A ask\n    class F care\n    class J done\n```\n\nThe shaded boxes are the two that matter: an empty result that **says** it is\nempty, and an output short enough to read to the end.\n\n```console\n$ ai-mem-search prisma\n240 match(es) for 'prisma'\npaths below /Users/you/_Ai_Memory\n--\n_session_logs/checkout-api/checkout-api-2026-08-31_00-00-00.md:56:  - **A green deploy log is not evidence a migration ran.** Verify against `_prisma_migrations`.\n...\n--\nshowing 25 of 240 (215 hidden), across 25 file(s), at most 1 line(s) each -- for more lines per file raise AI_MEM_SEARCH_PER_FILE, or narrow with: ai-mem-search 'prisma' <project>\n```\n\nUnder the hood it is one `grep -rniF`, a sort, and a cap. Each choice exists\nbecause the alternative gave a wrong answer on a real vault:\n\n| choice | why |\n|---|---|\n| **Case-insensitive** | A false empty is the worst answer a memory tool can give. `precompact` found 0 case-sensitively and 3 with `-i`; `Postgres` 56 vs 90. |\n| **Literal, not regex** | The caller is usually an agent passing free text, where a stray `.` or `(` must match itself. |\n| **Lessons first, then newest** | Session logs outnumber lessons ~4:1 and are always newer, so recency alone buried them — searching `snapshot` returned 25 logs and none of the 3 lessons answering it. A lesson now outranks newer chatter. |\n| **Bounded output** | Unbounded, a common term ran ~29k tokens and was silently truncated by the host — leaving the agent unable to tell *hidden* from *absent*. |\n| **One line per file** | Until the budget binds. Spends it on distinct notes rather than the chattiest one. |\n| **Compact lines** | Vault root printed once and stripped from every path; lines over 200 chars clamped. |\n| **One hop out** | Any `[[wikilink]]` on a hit resolves to its project note with a one-line excerpt, so a lesson tells you where it happened without a second search. |\n\n### What it costs, and what it cannot do\n\nMeasured, not estimated. Reproduce with `ai-mem-search <term>`.\n\n**Matching is substring, case-insensitive, literal** — not word matching, not\nstemming, not fuzzy:\n\n| query | matches | |\n|---|---|---|\n| `git` | 765 | substring — also hits `github` |\n| `github` | 255 | a subset of the above |\n| `Postgres` / `POSTGRES` / `postgre` | 93 each | case ignored, prefix matches |\n| `.env` | 131 | the dot is literal |\n| `env` | 307 | more — `.env` is narrower |\n| `postgress` *(typo)* | **0** | **no fuzzy matching** |\n\nThat last row is the limit: a misspelling finds nothing.\n\n**Search a few distinctive words.** The exact phrase is tried first. If it\nhas no hits, each word is tried on its own, and notes that hold more of the\nwords rank first, so word order does not decide the result: on a real vault\n`migration prisma` found **nothing** as a phrase, and now returns the Prisma\nmigration lessons first. Filler words (`the`, `how`, `not`…) are ignored. A\nbare tool name like `sed` alone returns over 2,000 lines, so add a word from\nthe symptom.\n\n**Token cost is bounded and flat:**\n\n| matches | tokens |\n|---|---|\n| 0 | ~23 |\n| 240 | ~1,224 |\n| 10,198 | ~1,270 |\n\nForty times the matches for the same cost.\n\n**Speed is linear in vault size**, dominated by how many lines match:\n\n| vault | rare term | common term |\n|---|---|---|\n| 501 notes *(real)* | — | **0.07 s** |\n| 2,000 *(synthetic)* | 0.35 s | 1.29 s |\n| 10,000 | 1.74 s | 7.20 s |\n\nComfortable to a few thousand notes — years of daily logs. Past ~10k, scope\nit: `ai-mem-search <term> <project>`. If that is not enough, this is the wrong\ntool and you want a real index.\n\n### Is it used while you code, or only at launch?\n\n**Pushed once at launch** (~4,300 tokens): your profile, standards, the\nproject note's path, a digest of the last session, and the lesson topics.\nThe last session is the newest log with content. A log that is still the\nunfilled template (a session opened and closed without notes) is skipped.\n\n**Pulled on demand after that.** Nothing re-injects. The agent reaches the\nvault mid-session only by running `ai-mem-search` itself — which the launch\nprompt tells it to do before solving anything, and again whenever it hits a\nblocker.\n\nSo: a reference library the agent is told to consult, not a memory it thinks\nwith. Two consequences — launch a bare `claude` instead of `claude-start` and\nyou get **no vault context at all**, and whether it helps mid-build depends on\nthe agent actually searching.\n\n### Mirrored notes are not injected twice\n\n`_Standards.md` ships declaring `mirror_of: _Global_Profile.md`. The launch\nprompt reads both, so injecting both in full restates text the model just\nread.\n\nWhen a note declares `mirror_of:`, only the differing lines are injected,\nunder a marker naming what was elided. On a real vault, 72 of the mirror's 80\nunique lines were verbatim duplicates: **launch prompt 7,200 → 4,300 tokens,\nnothing lost.**\n\nMatching is exact and line-by-line, so anything reworded survives. It fails\n**open** — no frontmatter, a missing source, or an empty diff all inject the\nnote in full. Injecting twice costs tokens; dropping a note costs the agent\ncontext it was promised.\n\n### About me: hard \"no\"s in full, the rest on demand\n\n`_about_me/` holds the human side of the profile as several small notes, not\none file. One file grows until the note cap cuts it off, like a long profile.\nEach note sets `inject:` in its frontmatter:\n\n| `inject:` | What reaches the prompt | Use it for |\n|---|---|---|\n| `always` | the body, at the **top** of the prompt, capped at 1,500 chars (`AI_MEM_ABOUT_ME_MAX_CHARS`) | non-negotiables |\n| anything else | one index line: the file name and its `read_when:` trigger | wants, interests, values, expertise, modes |\n\nHard \"no\"s are never behind a link: an agent that must open a file to see a\nrule often does not open it. The other notes cost one line each, however long\nthey grow, and the `read_when:` trigger tells the agent when to open them.\n\nBlockquotes and `[bracketed]` placeholder bullets are notes to you and are\nnever injected. A note that holds only those is not listed. A note whose name\nstarts with `_` is skipped. Write these notes yourself, or run `ai-about-me`\nto fill them by interview. An agent may propose a change, but a profile the\nagent edits is no longer your own view of yourself.\n\n### Every path into the prompt is bounded\n\nThe launch prompt is paid **every session, in every project**, so anything\nthat can grow without limit is a cost that compounds silently.\n\n| what reaches the prompt | bound |\n|---|---|\n| Profile and standards | **8,000 chars** each for shell launchers; GUI MCP is uncapped by default (`AI_MEM_NOTE_MAX_CHARS`) |\n| `_about_me/` | `inject: always` notes **1,500 chars** each (`AI_MEM_ABOUT_ME_MAX_CHARS`); every other note one index line |\n| Previous session outcome | 4 fields, 500 chars each |\n| Lesson index | 200 slugs, names only |\n| Project note | **a path, not its contents** |\n| `ai-mem-search` output | flat: 765 matches cost the same as 240 |\n\nThe profile was the one exception. It grew to 12,405 chars and pushed the\nlaunch prompt from 4,321 tokens to 5,125 without anything reporting it —\na 19% rise nobody asked for. Capping it brought it to **4,046**.\n\nTruncation always states the real total and where to read the rest:\n\n```text\n[truncated at 8000 of 12405 chars -- read the rest with: cat \"…/_Global_Profile.md\"]\n```\n\nThat line matters more than the saving. A truncated note the agent cannot\ndistinguish from a complete one is worse than a long one: it would answer\nfrom half a profile believing it had all of it. The cut also lands on a line\nboundary, because Markdown severed mid-line can leave half a heading or an\nunterminated code fence, which reads as content rather than as damage.\n\n### Lessons are indexed, not injected\n\n`_lessons/` entries are the only notes true across projects, and the hardest\nto find — searching for one means guessing a term for knowledge you don't\nknow you have. So the launch prompt lists **every lesson slug and no bodies**:\n\n```text\n- Lessons already recorded (96), listed by topic slug only. If one looks\n  relevant, read it with `ai-mem-search <slug>` before solving again.\n  prisma-connection-pool-exhaustion, decimal-money-precision-js, ...\n```\n\nA slug is the lesson compressed: `prisma-connection-pool-exhaustion` tells\nyou whether to open it without opening it. ~8-10 tokens each, so the index\ngrows slowly: 96 lessons cost ~830, and a vault grown to 182 lessons costs\n~1,740 (`ai-context | grep -A3 'Lessons already recorded' | wc -c`, chars divided\nby ~4).\n\nNewest first, capped at `AI_MEM_LESSON_INDEX_LIMIT` (200). The cap matters —\npast a few hundred, most titles are irrelevant to any given session, and a\nlong tail of near-misses is the most damaging kind of distractor for a model.\n\nReturns non-zero on no matches, a missing term, or an unknown project, so\nit composes in a script.\n\n## Memory hygiene\n\nA second brain that only grows, rots. Old session logs bury the recent ones,\nnear-duplicate lessons pile up, and a dead `[[link]]` points at a note that is\nno longer there. Search then spends more tokens to return worse results.\n\ncreate-ai-memory keeps the vault healthy with the two-sided strategy a brain\nuses: it strengthens what gets used and lets the rest fade, but it never throws\naway what was hard to learn.\n\n| move | what it does | command |\n|---|---|---|\n| **Strengthen** | A lesson recalled again ranks higher, not just newer -- long-term potentiation for the vault. | `ai-mem-search` ranking |\n| **Consolidate** | Many session logs about one thing distil into a single durable lesson. | `ai-lesson` |\n| **Decay** | Old session logs move out of the hot search path into `_archive/`, still on disk and in git. Unfilled logs go after a day. | `ai-mem-sleep` |\n| **Prune (detect)** | Links to notes not written yet are ranked by how many notes link to them, never auto-removed. | `ai-mem-lint` |\n| **Schedule** | Run the whole pass nightly on its own. | `ai-mem-sleep-schedule` |\n\n**One rule is inviolable: a lesson is never pruned.** The failure-to-fix records\nin `_lessons/` are the durable knowledge -- the answer you may need once every\nsix months when an old blocker returns -- and pruning by disuse would delete\nexactly those. Only the ephemeral layer (session logs) decays, and even it is\narchived, not deleted. Reinforcement acts on lessons; decay acts on logs; the\ntwo never cross.\n\n### Where the model comes from\n\nThe design is modeled on how a brain manages memory. It is a design analogy,\nnot a claim to reproduce the biology. Two ideas guided it:\n\n- **Synaptic pruning and homeostasis.** An infant brain overproduces neural\n  connections, then removes the weak or unused ones from experience -- \"use it\n  or lose it\" -- which is what keeps it efficient. A vault that only appends is\n  the brain that never prunes.\n- **Long-term potentiation (LTP).** A pathway that fires repeatedly is\n  strengthened. That is the half a naive \"drop the old\" policy forgets, and the\n  half `ai-mem-search`'s reinforcement adds.\n\nFraming these as one scheduled maintenance pass -- consolidation, decay,\ngarbage collection, and graph pruning, run while the agent \"sleeps\" -- follows\nAndrew Orobator, [\"Your Agent Needs a Bedtime\"][bedtime], and the wider NeuroAI\nwork that applies synaptic pruning to artificial networks. Its rule that error\nrecoveries are never pruned is the same rule stated above.\n\n[bedtime]: https://medium.com/@andreworobator/your-agent-needs-a-bedtime-cc594be3f993\n\n## Session skills (optional)\n\nYou define your own per-session skills; create-ai-memory ships none. At launch it asks\ny/n for each skill you registered, then injects the chosen instruction blocks into\nthe agent (and for Cursor, writes them as a managed always-apply rule). Register\nnothing and every session is plain.\n\nAdd skills in `~/.zshrc` before sourcing the module. Each entry maps a `key` to\n`prompt::instruction block`, and `AI_MEM_SKILL_ORDER` sets the ask order:\n\n```zsh\ntypeset -gA AI_MEM_SKILLS\nAI_MEM_SKILLS[terse]='Use terse output this session?::Respond tersely; drop filler and hedging.'\nAI_MEM_SKILLS[design]='Use strict UI design discipline?::Apply careful frontend/UI design review to any design work.'\nAI_MEM_SKILL_ORDER=(terse design)\n\nsource \"$HOME/ai-memory/shell/ai-mem.zsh\"\n```\n\nThe injected block applies for the whole run, so a session's chosen skills persist\nas instructions across every change the agent makes.\n\n### Recommended skills to wire in\n\nThese are the session skills worth having on the picker. Each maps to a working\nClaude Code skill; if you have the skill installed, the block below tells the agent\nto use it, and the behavior still applies to agents that do not have the skill\nbecause the instruction is inlined. Drop them into `AI_MEM_SKILLS` and reorder to\ntaste.\n\n```zsh\ntypeset -gA AI_MEM_SKILLS\n\n# caveman: terse, no-filler output\nAI_MEM_SKILLS[caveman]='Terse output (caveman) this session?::Respond terse like a smart caveman. Keep technical substance exact; drop filler, pleasantries, and hedging.'\n\n# ponytail: the laziest solution that actually works (stdlib > deps, one line > fifty)\nAI_MEM_SKILLS[ponytail]='Minimal-code (ponytail) this session?::Use the ponytail approach. Prefer the standard library before new code, native platform features before dependencies, and one line before fifty. Question whether the task needs to exist at all (YAGNI).'\n\n# hallmark: anti-AI-slop UI and frontend design\nAI_MEM_SKILLS[hallmark]='Strict UI design (hallmark) this session?::Use the hallmark approach for any frontend, UI, or design work. Avoid generic AI-slop layouts; make type, spacing, color, and hierarchy intentional.'\n\nAI_MEM_SKILL_ORDER=(caveman ponytail hallmark)\n\nsource \"$HOME/ai-memory/shell/ai-mem.zsh\"\n```\n\n| Skill | What it does | Reach for it when | Source |\n|---|---|---|---|\n| **caveman** | Strips output to terse, no-filler answers | You want signal over prose | [caveman.so](https://caveman.so/) |\n| **ponytail** | Pushes the smallest change that works; stdlib and native features over dependencies | Building features or reviewing for over-engineering | [DietrichGebert/ponytail](https://github.com/DietrichGebert/ponytail) |\n| **hallmark** | Anti-AI-slop design discipline for UI and frontend work | Any visual or frontend task | [usehallmark.com](https://www.usehallmark.com/) |\n\nSkills are per session and independent, so you can turn on `ponytail` for a\nrefactor and add `hallmark` only when you touch the UI. These are separate,\ninstallable Claude Code skills; the blocks above inline their behavior so a\nsession still benefits even on an agent that does not have the skill installed.\n\n## Your vault\n\nThe vault is a folder of Markdown. Point `AI_MEM_ROOT` at a new directory or at an\nexisting Obsidian vault; either way you get graph view, backlinks, and full-text\nsearch over your own AI memory, plus plain `grep` when you want it.\n\n```\n$AI_MEM_ROOT/\n  _Global_Profile.md          your cross-project rules      (injected every session)\n  _Standards.md               extra shared standards        (injected every session)\n  _about_me/\n    non-negotiables.md         hard \"no\"s                    (injected every session)\n    wants.md, interests.md,    who you are                   (one index line each;\n    values.md, expertise.md,                                  read on demand)\n    modes.md\n  _projects/\n    _project_template.md       scaffold for new project notes\n    <repo>.md                  per-project durable context\n  _session_logs/\n    _session_template.md       scaffold for new session logs\n    <repo>/\n      <repo>-<timestamp>.md    one file per session\n  _lessons/\n    _lesson_template.md       scaffold for new lesson topics\n    <topic-slug>.md            cross-project decisions/mistakes, filed by ai-lesson\n  _transcripts/\n    <source-id>.md             distilled meetings, calls, videos and voice notes,\n                               filed by ai-mem-ingest\n```\n\nNotes are created from templates on first use and never overwritten. Edit\n`_Global_Profile.md`, `_Standards.md` and the notes in `_about_me/` to make them\nyours; the shipped versions are sanitized placeholders. `_about_me/` is seeded\nonly when the folder does not exist, so a note you delete stays deleted.\n\n## Open Knowledge Format\n\n[OKF](https://github.com/GoogleCloudPlatform/open-knowledge-format) is\nGoogle's vendor-neutral spec for portable, agent-readable knowledge. A bundle\nis just a directory tree of markdown files — shippable as a repo, tarball or\nzip, readable by any agent, Obsidian, MkDocs or graph viewer without a\ntranslation layer.\n\nYour vault is one, by convergence rather than adoption. The\n[spec](https://github.com/GoogleCloudPlatform/knowledge-catalog/blob/main/okf/SPEC.md)\nrequires no file at the bundle root and exactly one frontmatter key:\n*\"`type` is the only always-required key.\"*\n\n| OKF | Here |\n|---|---|\n| Markdown in a directory hierarchy | ✅ |\n| YAML frontmatter | ✅ every note |\n| **`type` — the only required key** | ✅ `ai-project-context`, `ai-lesson`, `ai-session-log`, … |\n| Recommended `title` / `description` / `tags` | ⚠️ partial |\n| Cross-links as markdown links | ⚠️ `[[wikilink]]` instead |\n| `index.md` / `log.md` | optional in the spec; not generated |\n| No SDK, no account, git-versionable | ✅ |\n\nOne real deviation: **links stay Obsidian-style**, because the graph view and\nbacklinks are much of why anyone keeps a vault. No `index.md` is generated —\n`grep` never reads one and a materialized index goes stale on rename.\n\nSession logs predate the `type` field, so a vault in use holds notes without\nit:\n\n```console\n$ ai-mem-lint --fix\nbackfilled `type: ai-session-log` into 350 session log(s)\n```\n\nEdits in place, preserves frontmatter and body, idempotent.\n\n## Add another harness\n\nLaunchers are configurable. Each harness is one small adapter, and the\n`<name>-start` function is generated for you. `claude`, `codex`, `agy`,\n`gemini`, `cursor`, `opencode`, and `omp` ship built in.\n\n`agy` is the [Antigravity CLI](https://antigravity.google), which in practice\nhas replaced the Gemini CLI. Both adapters ship: `gemini-start` still works\nfor anyone using it, and `agy-start` is the one to reach for now.\n\n`omp-start` passes the vault as an OMP workspace directory and adds the selected\nsession skills through OMP's system-prompt channel. It also tells OMP to use\nand prints the active OMP provider/model selector plus a redacted\nprovider-usage snapshot before the TUI opens. The snapshot includes\nweekly/account limits when the provider exposes them. Set\n`AI_MEM_OMP_SHOW_USAGE=0` to skip that preflight check. Plain `omp` is\nunchanged; it does not load the ai-memory vault or session context.\n\nGUI clients that need OMP RPC or ACP can still invoke the literal `omp-start`\npipeline. Set `AI_MEM_NONINTERACTIVE=1`, then run `omp-start --mode rpc` or\n`omp-start acp`. The launcher skips the terminal skill picker and moves the\nassembled vault context into OMP's system-prompt channel before the first user\nturn. Set `AI_MEM_MANUAL_SKILLS` to explicit pipe-separated skill keys when a\nnoninteractive client must enable skills automatically without asking.\n\nTo add `aider`:\n\n1. Define the adapter in `shell/adapters.zsh`. It receives `$1` memory prompt,\n   `$2` mode block, and `$3` onward extra args:\n   ```zsh\n   __ai_adapter_aider() {\n       local memory_prompt=\"$1\"\n       aider --message \"$memory_prompt\"\n   }\n   ```\n2. Register it in `~/.zshrc` before sourcing, or edit the default:\n   ```zsh\n   export AI_MEM_HARNESSES=\"claude codex agy gemini cursor opencode omp aider\"\n   ```\n3. `aider-start` now exists. No core edits.\n\n`cursor-start` targets `cursor-agent`, Cursor's CLI, which takes the prompt\npositionally. If only the `cursor` GUI is installed it falls back to opening\nthat instead -- the GUI has no prompt path, so the session's skills still\narrive via the managed rule file but the vault context does not.\n\nName private helpers with **two** leading underscores (`__ai_adapter_aider`,\nnot `_ai_adapter_aider`). Claude Code replays a snapshot of your interactive\nshell for every command it runs, and that snapshot drops single-underscore\nfunction names -- the filter targets zsh's completion functions, but it takes\nprivate helpers with it. A one-underscore helper simply will not exist inside\nan agent-run command.\n\n## Integrations\n\n### Graph view\n\n`ai-mem-search` answers a question you already knew to ask. This answers the\nother one — what is in there, and what turned out to be connected.\n\nAsk the agent (\"open my memory\") or run it:\n\n```sh\nai-mem-serve          # opens a browser at http://127.0.0.1:7777\n```\n\n![The vault as a graph: notes coloured by type, a lesson selected, its linked project highlighted](assets/graph-view.png)\n\nNodes are coloured by `type` and sized by how many notes link to them.\nSelecting one dims everything unconnected and renders the note beside the\ngraph. *Durable knowledge* is the default view — session logs usually\noutnumber everything else and bury the rest. A note carrying a dangling\n`[[link]]` wears a dashed red ring, so a dead edge is visible in the graph as\nwell as in the note (where it renders dotted and unclickable).\n\nLoopback only, deliberately: the vault holds project history. Zero\ndependencies — the layout is a small force simulation, not a charting\nlibrary, so it works offline. Run it twice and the second call just opens the\ntab, unless another vault holds the port.\n\n### GUI clients (MCP)\n\nThe launchers reach an agent through its opening prompt. **A GUI opened from\nthe Dock never runs one**, so Claude Desktop and the Cursor GUI would see\nnothing. `ai-mem-mcp` is their channel.\n\n**Claude Desktop** — `~/Library/Application Support/Claude/claude_desktop_config.json`\n(**Cursor** — `~/.cursor/mcp.json`, same shape):\n\n```json\n{\n  \"mcpServers\": {\n    \"ai-memory\": {\n      \"command\": \"ai-mem-mcp\",\n      \"env\": { \"AI_MEM_ROOT\": \"/absolute/path/to/your/_Ai_Memory\" }\n    }\n  }\n}\n```\n\nSix tools, read **and** write, so a GUI session is not a dead end:\n`search_memory`, `get_context`, `read_note`, `add_note`, `add_lesson`,\n`open_graph`. Each shells out to the same zsh functions the CLI uses — one\nimplementation, not a copy that drifts. A note written from Claude Desktop is\ncommitted and pushed exactly like one written in a terminal.\n\nThe MCP initialization response carries the global profile and standards, plus\na standing brief (read context first, search before solving, write back after)\n-- without it a model has no reason to suspect a memory exists. Tool\ndescriptions stay terse because schemas are re-sent every turn.\n\nA GUI session log is created by the first write, not by a read. Each MCP\nserver keeps one log path for its lifetime; `add_note` (or an agent that\nwrites the session outcome) creates the file there. A conversation that only\nreads context leaves no empty note in the vault.\n\n**Do not register this for a terminal agent.** Claude Code, Codex, Gemini and\nopencode have a shell and should call `ai-mem-search` directly.\n\n**A desktop app that has a shell but no launcher** (the Codex app, for\nexample) gets no vault context, because nothing runs `codex-start` for it.\nWithout instructions, it guesses: it takes the newest log in the vault, which\ncan be a different project. Put the rules in the app's global instruction file\n(`~/.codex/AGENTS.md` for Codex). A GUI app does not read `~/.zshrc`, so give\nthe vault path and the module path explicitly:\n\n```md\n## AI memory vault\nBefore you answer a question about a project, load its context:\n`AI_MEM_ROOT=/abs/path/_Ai_Memory zsh -c 'source /abs/path/create-ai-memory/shell/ai-mem.zsh; ai-context <project>'`\n- <project> is the name the user gives, or the git repo folder name.\n- If the user gives no project and the folder is not a git repo, STOP and ask.\n  A folder that is not a repo does not give a project name. Do not guess.\n- If `_projects/<project>.md` does not exist, ask before you continue:\n  `ai-context` makes a new empty project for a name that is not correct.\n- Skip this step if the prompt already contains vault context.\n- To record a note, use the same form with `AI_MEM_ACTIVE_PROJECT=<project>` added\n  and `ai-note \"<text>\"`. Outside a repo, that variable names the project.\n  Each shell command starts a new process, so set it on every call.\n```\n\n### Claude Code hooks\nThe files live in `hooks/claude/`. Record repo `HEAD` at session start, then on\nexit write an auto block to the log with the branch, commits made this session, and\nuncommitted changes, so the next session has real carryover instead of an empty\ntemplate. A third hook fires right before Claude Code auto-compacts: the first\nattempt each session it blocks compaction once and tells the agent to run\n`ai-note` before its working context gets thrown away, then gets out of the\nway for every later attempt (one-shot, via a marker file next to the log) so\nthe session can never get stuck refusing to compact. It only guards\nauto-compact, not a deliberate `/compact`. Merge `settings.snippet.json` into\n`~/.claude/settings.json`, replacing `<AI_MEM_HOME>` with an absolute path.\nAll three hooks no-op for plain `claude` runs; they gate on\n`$AI_MEM_ACTIVE_SESSION_LOG`. That variable stays exported in your shell after\nthe agent exits, so the session-start and Stop hooks also check\n`$AI_MEM_ACTIVE_GIT_DIR`: they write only in the repo the session was launched\nin, or a worktree of it. A later plain `claude` in a different repo does not\nput that repo's commits in the wrong project's log. This is Claude-Code-specific -- other agents\n(Codex, Gemini, etc.) have no equivalent pre-compaction hook, so jotting\nthings down with `ai-note`/`codex-note` as you go still matters there.\n\n### Git commit guard\nThe files live in `hooks/git/`:\n\n- **`commit-msg`** requires Conventional Commits and a structured body, and that\n  `ai-context` was loaded in the committing shell, matching a per-repo token. An\n  agent cannot commit without the vault context loaded.\n- **`pre-push`** blocks direct pushes to `main` unless `ALLOW_PUSH_TO_MAIN=1`.\n\nEnable per repo:\n\n```sh\ncp ~/ai-memory/hooks/git/* <repo>/.githooks/\ngit -C <repo> config core.hooksPath .githooks\n```\n\n## Configuration\n\n| Env var | Default | Purpose |\n|---|---|---|\n| `AI_MEM_ROOT` | `$HOME/.ai-memory/_Ai_Memory` | Vault root. Point at any folder, including an existing Obsidian vault |\n| `AI_MEM_HARNESSES` | `claude codex agy gemini cursor opencode omp` | Space-separated harnesses to generate `-start` functions for |\n| `AI_MEM_AGENTS` | compatibility alias | Older installs may keep using this name; `AI_MEM_HARNESSES` takes precedence |\n| `AI_MEM_PREFERRED_HARNESS` | empty | Optional override used by every `*-start`; e.g. `omp` makes `codex-start` launch OMP while keeping the familiar entry point |\n| `AI_MEM_OMP_SHOW_USAGE` | `1` | Show OMP's redacted provider/account usage snapshot before an ai-memory OMP session; set to `0` to skip |\n| `AI_MEM_SKILLS` / `AI_MEM_SKILL_ORDER` | empty | Your per-session skills (see above) |\n| `AI_MEM_NONINTERACTIVE` | empty | Skip the terminal skill picker while keeping the literal `*-start` session pipeline; intended for GUI RPC/ACP launchers |\n| `AI_MEM_MANUAL_SKILLS` | empty | Pipe-separated skill keys to inject without the picker; also supplies skills for noninteractive launchers |\n| `AI_MEM_LESSON_INDEX_LIMIT` | `200` | Lesson slugs listed in the launch prompt before it truncates to the newest. Names only -- bodies are never injected |\n| `AI_MEM_NOTE_MAX_CHARS` | `8000` for shell launchers; `0` for GUI MCP | Cap on what one inlined note contributes to the launch prompt. Truncation states the real total and the path. `0` disables it |\n| `AI_MEM_ABOUT_ME_MAX_CHARS` | `1500` | Cap on each `inject: always` note in `_about_me/`. Kept small on purpose: these notes are paid for in every session. Truncation states the real total and the path |\n| `AI_MEM_SEARCH_LIMIT` | `25` | Result lines `ai-mem-search` prints before it truncates. The default is sized for an agent's context window; raise it when you are reading the output yourself (see [How search works](#how-search-works)) |\n| `AI_MEM_HOME` | the module's own directory | Exported by the module when it is sourced, not set by you. A cron job, a launchd agent or any non-interactive shell can re-source the module with `zsh -c 'source \"$AI_MEM_HOME/ai-mem.zsh\"; ...'` instead of a hard-coded path |\n| `AI_MEM_SEARCH_PER_FILE` | `1` | Lines shown per file once the cap binds. Spreads results across notes instead of on the chattiest one; ignored when every match already fits |\n\n## Why not just CLAUDE.md?\n\nYou probably already have one. Keep it — this does a different job, and the\ntool defers to it by design.\n\n| | `CLAUDE.md` / `AGENTS.md` | create-ai-memory |\n|---|---|---|\n| Holds | rules and conventions | what happened, decided, and broke |\n| Changes | when you edit it | every session, automatically |\n| Scope | that one repo | global + per-repo + cross-project lessons |\n| Agents | one vendor's file per agent | one vault, five CLIs |\n| Answers | *\"how should you work here?\"* | *\"what did we already try?\"* |\n\nA `CLAUDE.md` is a standing instruction. It does not know that last Tuesday\nyou found the connection pool was sized below the worker count, or that the\nsame bug bit a different repo in March. That is what accumulates here.\n\nThey compose: the injected profile explicitly tells the agent that repo-local\ninstruction files **override it** for that repo. Nothing to migrate, nothing\nto delete.\n\n## Troubleshooting\n\n**`claude-start` works but the agent has no memory of anything.**\nYou launched plain `claude`. Nothing hooks it — the vault reaches the agent\nthrough the launcher's opening prompt, so a bare CLI gets none of it.\n\n**A command \"isn't installed\" right after an upgrade.**\nYour shell is stale. These are zsh functions loaded when the terminal opened;\nediting or upgrading the module changes nothing in a shell already running.\nOpen a new terminal, or `exec zsh`. `which` under bash will not find them\neither — use `command -v`.\n\n**The agent's \"Project context\" is bracket placeholders.**\nThe project note was never filled in. The launch prompt now says so and asks\nthe agent to fill it from the repo; you can also edit\n`_projects/<repo>.md` yourself. It is read at every launch for that repo, so\nit is worth ten minutes once.\n\n**`ai-mem-lint` reports orphaned session logs.**\nTheir `project:` frontmatter is not a `[[wikilink]]`, usually because the\nvault's session template predates the current one. Scaffolding only fills in\n*missing* files, so an existing vault keeps its old templates. Run\n`ai-mem-lint --fix`.\n\n**Search returns nothing for a term you are sure you wrote.**\nWords match as literal substrings with no fuzzy matching, so a typo finds\nzero. Use a few distinctive words rather than a whole error line — see\n[what it costs](#what-it-costs-and-what-it-cannot-do).\n\n## Why plain files\n\ncreate-ai-memory is deliberately small. There is no server to run, no container to pull,\nno database to migrate, no API key to store. The design choices behind that:\n\n- **The vault is the source of truth; the chat is disposable.** Durable state\n  lives in Markdown you can read, diff, and version, not inside any agent.\n- **Files over a service.** A folder syncs over git, Dropbox, or Syncthing, opens\n  in Obsidian, and greps in a shell. It outlives any one tool.\n- **Explicit over magic.** You decide what goes in the profile, the project note,\n  and the session log. The agent reads them; it does not silently rewrite your\n  memory behind your back.\n- **Path-guarded writes.** Every file operation is checked to stay inside\n  `$AI_MEM_ROOT`, so an agent cannot write outside the memory boundary.\n- **Project equals git repo.** Resolution prefers the repo you are standing in, so\n  moving between projects in one shell never pins the wrong project.\n\nIf you want an auto-capturing server with a web UI and vector search, other tools\ndo that. create-ai-memory trades those for something you can read end to end in an\nafternoon and carry anywhere.\n\n## Tests\n\nOffline unit suite (throwaway vault and git repo, no network): path guarding,\nproject resolution, session prep, the context prompt, the skill picker, launcher\ngeneration, adapter dispatch, the commit token, `ai-note`, `ai-mem-ingest`, and\nthe cursor rule file.\n\n```sh\nzsh tests/run.sh     # offline unit tests (340 assertions)\nzsh tests/smoke.sh   # live: launches each agent headlessly, checks it responds\n```\n\n`smoke.sh` makes real API calls, so each CLI must be installed and authed\n(opencode defaults to DeepSeek; set it up or pass\n`AIMEM_SMOKE_OPENCODE_MODEL=provider/model`).\n\n## FAQ\n\n**Does it send my code anywhere?**\nNo. create-ai-memory is shell functions plus Markdown files on your disk. The only network\ncalls are the ones your agent already makes.\n\n**Do I need Obsidian?**\nNo. The vault is plain Markdown. Obsidian is a nice way to browse it, not a\nrequirement.\n\n**Do I need an API key or a paid plan?**\nNo. create-ai-memory itself needs neither. Your agents use whatever auth they already have.\n\n**Which shells and platforms?**\nzsh today, on macOS and Linux (including WSL). A bash port is on the roadmap.\n\n**Does plain `claude` still work?**\nYes. Only the `*-start` launchers inject memory; plain runs are untouched, and the\nhooks no-op unless a session was launched through create-ai-memory.\n\n**Can I use my existing Obsidian vault?**\nYes. Point `AI_MEM_ROOT` at it. Notes are additive and never overwrite your files.\n\n**How is a \"project\" identified?**\nBy the directory name of the git repo you are in.\n\n**My agent isn't listed. Can I add it?**\nYes, if it has a CLI. See [Add another agent](#add-another-agent); it is three\nlines.\n\n## Roadmap\n\n- Bash port of the shell module.\n- More agent adapters shipped by default: aider and others.\n- Optional cross-project index and search over session logs.\n\n## Contributing\n\nClone the repo and run `install.sh` from the checkout — that's the source of\ntruth every other install path (npm, plugin managers) reuses:\n\n```sh\ngit clone https://github.com/rambaarde/create-ai-memory.git ~/ai-memory\n~/ai-memory/install.sh\n```\n\nIssues and PRs are welcome. Good first areas: a bash port, new agent adapters, and\ndocs. Run `zsh tests/run.sh` before opening a PR; keep changes additive and the\nvault path-guarded.\n\nPersistent AI memory should not be a personal hack; it should be something the\nwhole community can install.\n\n## License\n\nMIT © Ram Christopher Baarde\n","readmeFilename":"README.md"}