{"_id":"zerodb-sequential-thinking-mcp","_rev":"3-b70fe0700c84a938c0f141d14638568a","name":"zerodb-sequential-thinking-mcp","dist-tags":{"latest":"0.1.2"},"versions":{"0.1.0":{"name":"zerodb-sequential-thinking-mcp","version":"0.1.0","keywords":["mcp","mcp-server","mcp-server-hosting","model-context-protocol","model-context-protocol-server","sequential-thinking","sequential-reasoning","chain-of-thought","chain-of-thought-reasoning","reasoning","multi-step-reasoning","problem-solving","structured-thinking","deep-thinking","self-provisioning","agent-provisioning","zero-config","auto-provisioning","zero-signup","instant-setup","agent-onboarding","agent-memory","persistent-memory","cross-session","cross-session-memory","cross-session-reasoning","plan-artifacts","context-management","context-window","conversation-history","long-term-memory","zerodb","ainative","openclaw","openclaw-ai","agent-swarm","zerolocal","claude","claude-code","claude-desktop","claude-ai-agents","codex","openai-agents","openai-codex","windsurf","cursor","cline","continue","vscode","langchain","crewai","autogen","n8n-ai","multi-agent","multi-agent-architecture","agent-orchestration","llm","ai","agentic-ai","ai-agents","ai-tool-calling","rag","retrieval-augmented-generation","local-first-ai"],"author":{"url":"https://ainative.studio","name":"AINative Studio","email":"toby@rely.ventures"},"license":"MIT","_id":"zerodb-sequential-thinking-mcp@0.1.0","maintainers":[{"name":"ainative-studio","email":"toby@rely.ventures"}],"homepage":"https://github.com/AINative-Studio/zerodb-sequential-thinking-mcp#readme","bugs":{"url":"https://github.com/AINative-Studio/zerodb-sequential-thinking-mcp/issues"},"bin":{"zerodb-sequential-thinking-mcp":"index.js"},"dist":{"shasum":"d7e44633e50acf85cadfdfe0e4ccf09deb337d01","tarball":"https://registry.npmjs.org/zerodb-sequential-thinking-mcp/-/zerodb-sequential-thinking-mcp-0.1.0.tgz","fileCount":11,"integrity":"sha512-bxtUwB/YCvFP5pfGAWpB1q+dDFmYKVInhVEHl3WW3B0BFPi5aEfrrQB98g6xvNp+MBfLCTM46pKWsYsxCS9ZVA==","signatures":[{"sig":"MEUCIQC309k2LBrj47/jcbaQzd9Jay8M0Ya9gxZZJolQfBHg2wIgXlClGlg/ZPtuWl+I14vzqAEyXnoX2TVFHjuOXFlflqk=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"unpackedSize":26771},"main":"index.js","type":"module","engines":{"node":">=18.0.0"},"gitHead":"5298ea26237ff459ff97bba76a51815d3c05bbc5","scripts":{"start":"node index.js"},"_npmUser":{"name":"ainative-studio","email":"toby@rely.ventures"},"repository":{"url":"git+https://github.com/AINative-Studio/zerodb-sequential-thinking-mcp.git","type":"git"},"_npmVersion":"10.9.4","description":"Persistent sequential thinking MCP — chain-of-thought reasoning that survives sessions, resumes across agents, and saves conclusions as plan artifacts. Powered by ZeroDB. Zero-config: agents self-provision a ZeroDB account on first run, no signup required","directories":{},"_nodeVersion":"22.21.0","dependencies":{"dotenv":"^16.4.0","@modelcontextprotocol/sdk":"^0.5.0"},"publishConfig":{"access":"public","registry":"https://registry.npmjs.org/"},"_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/zerodb-sequential-thinking-mcp_0.1.0_1779000758794_0.5652529260364263","host":"s3://npm-registry-packages-npm-production"}},"0.1.1":{"name":"zerodb-sequential-thinking-mcp","version":"0.1.1","keywords":["mcp","mcp-server","mcp-server-hosting","model-context-protocol","model-context-protocol-server","sequential-thinking","sequential-reasoning","chain-of-thought","chain-of-thought-reasoning","reasoning","multi-step-reasoning","problem-solving","structured-thinking","deep-thinking","self-provisioning","agent-provisioning","zero-config","auto-provisioning","zero-signup","instant-setup","agent-onboarding","agent-memory","persistent-memory","cross-session","cross-session-memory","cross-session-reasoning","plan-artifacts","context-management","context-window","conversation-history","long-term-memory","zerodb","ainative","openclaw","openclaw-ai","agent-swarm","zerolocal","claude","claude-code","claude-desktop","claude-ai-agents","codex","openai-agents","openai-codex","windsurf","cursor","cline","continue","vscode","langchain","crewai","autogen","n8n-ai","multi-agent","multi-agent-architecture","agent-orchestration","llm","ai","agentic-ai","ai-agents","ai-tool-calling","rag","retrieval-augmented-generation","local-first-ai"],"author":{"url":"https://ainative.studio","name":"AINative Studio","email":"toby@rely.ventures"},"license":"MIT","_id":"zerodb-sequential-thinking-mcp@0.1.1","maintainers":[{"name":"ainative-studio","email":"toby@rely.ventures"}],"homepage":"https://github.com/AINative-Studio/zerodb-sequential-thinking-mcp#readme","bugs":{"url":"https://github.com/AINative-Studio/zerodb-sequential-thinking-mcp/issues"},"bin":{"zerodb-sequential-thinking-mcp":"index.js"},"dist":{"shasum":"e5cdc8c20da87539533a1937f1731a4ab0372efb","tarball":"https://registry.npmjs.org/zerodb-sequential-thinking-mcp/-/zerodb-sequential-thinking-mcp-0.1.1.tgz","fileCount":11,"integrity":"sha512-gBSeOvwXwyLwflJbOvl4F7dFiYQ5p9On17Q0Ih5RjQGDazoN5Ieq9C2AqtcbW+2cFtJasxIvD+Og1BLuukxyVQ==","signatures":[{"sig":"MEQCIE0LyysSOAAvgsPBHZsKBbAHP4F8l+WjVQl3i4d4LRfUAiAuVedDCkc4CeS/9s8xVWa8/O5l7rSsBeF0YCiYEXP3mQ==","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"unpackedSize":27280},"main":"index.js","type":"module","engines":{"node":">=18.0.0"},"gitHead":"4f7ed77b25fe69b26d86917b15a9bb6bd034080f","scripts":{"start":"node index.js"},"_npmUser":{"name":"ainative-studio","email":"toby@rely.ventures"},"repository":{"url":"git+https://github.com/AINative-Studio/zerodb-sequential-thinking-mcp.git","type":"git"},"_npmVersion":"10.9.4","description":"Persistent sequential thinking MCP — chain-of-thought reasoning that survives sessions, resumes across agents, and saves conclusions as plan artifacts. Powered by ZeroDB. Zero-config: agents self-provision a ZeroDB account on first run, no signup required","directories":{},"_nodeVersion":"22.21.0","dependencies":{"dotenv":"^16.4.0","@modelcontextprotocol/sdk":"^0.5.0"},"publishConfig":{"access":"public","registry":"https://registry.npmjs.org/"},"_hasShrinkwrap":false,"_npmOperationalInternal":{"tmp":"tmp/zerodb-sequential-thinking-mcp_0.1.1_1779000989075_0.8233158793032311","host":"s3://npm-registry-packages-npm-production"}},"0.1.2":{"_id":"zerodb-sequential-thinking-mcp@0.1.2","bin":{"zerodb-sequential-thinking-mcp":"index.js"},"bugs":{"url":"https://github.com/AINative-Studio/zerodb-sequential-thinking-mcp/issues"},"dist":{"shasum":"0e6a2093fd0a65945916ade2e41d18eb05546c27","tarball":"https://registry.npmjs.org/zerodb-sequential-thinking-mcp/-/zerodb-sequential-thinking-mcp-0.1.2.tgz","fileCount":11,"integrity":"sha512-/Bnvwe2IOyz951c0ihyFkqzzVn0GI8d4ti5wQoolqFR2/1Xi56L9eMB3eCRk2osueyfN80dB67DqaSc8sYovQw==","signatures":[{"sig":"MEUCIEpltJeHrPC0uUphOSD9HdDBiprJy4RgGidneAkq9nWyAiEA5RhZJskoM2BdzkBRIKo8ICBKubn6BO2yTaaxF5tXDuc=","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"},{"keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U","sig":"MEQCID8R6LZ8iy9modKsqu9yzHkP8UnZXJsP8npuaFd19BHlAiAq4e37n/tgYe3MF6sk7LFtlerFiypzTVn+QSb44GMfdw=="}],"unpackedSize":27247},"main":"index.js","name":"zerodb-sequential-thinking-mcp","type":"module","author":{"url":"https://ainative.studio","name":"AINative Studio","email":"toby@rely.ventures"},"engines":{"node":">=18.0.0"},"gitHead":"765152cc6b7c25008f4cd275bd0cb26415cc4397","license":"MIT","scripts":{"start":"node index.js"},"version":"0.1.2","_npmUser":{"name":"ainative-studio","email":"toby@rely.ventures"},"homepage":"https://ainative.studio/products/zerodb","keywords":["mcp","mcp-server","mcp-server-hosting","model-context-protocol","model-context-protocol-server","sequential-thinking","sequential-reasoning","chain-of-thought","chain-of-thought-reasoning","reasoning","multi-step-reasoning","problem-solving","structured-thinking","deep-thinking","self-provisioning","agent-provisioning","zero-config","auto-provisioning","zero-signup","instant-setup","agent-onboarding","agent-memory","persistent-memory","cross-session","cross-session-memory","cross-session-reasoning","plan-artifacts","context-management","context-window","conversation-history","long-term-memory","zerodb","ainative","openclaw","openclaw-ai","agent-swarm","zerolocal","claude","claude-code","claude-desktop","claude-ai-agents","codex","openai-agents","openai-codex","windsurf","cursor","cline","continue","vscode","langchain","crewai","autogen","n8n-ai","multi-agent","multi-agent-architecture","agent-orchestration","llm","ai","agentic-ai","ai-agents","ai-tool-calling","rag","retrieval-augmented-generation","local-first-ai"],"repository":{"url":"git+https://github.com/AINative-Studio/zerodb-sequential-thinking-mcp.git","type":"git"},"_npmVersion":"10.9.4","description":"Persistent sequential thinking MCP — chain-of-thought reasoning that survives sessions, resumes across agents, and saves conclusions as plan artifacts. Powered by ZeroDB. Zero-config: agents self-provision a ZeroDB account on first run, no signup required","directories":{},"maintainers":[{"name":"ainative-studio","email":"toby@rely.ventures"}],"_nodeVersion":"22.21.0","dependencies":{"dotenv":"^16.4.0","@modelcontextprotocol/sdk":"^0.5.0"},"publishConfig":{"access":"public","registry":"https://registry.npmjs.org/"},"_hasShrinkwrap":false,"_npmOperationalInternal":{"host":"s3://npm-registry-packages-npm-production","tmp":"tmp/zerodb-sequential-thinking-mcp_0.1.2_1790310706371_0.9047273900585031"}}},"time":{"created":"2026-05-17T06:52:38.742Z","modified":"2026-09-25T04:31:46.603Z","0.1.0":"2026-05-17T06:52:38.971Z","0.1.1":"2026-05-17T06:56:29.217Z","0.1.2":"2026-09-25T04:31:46.464Z"},"bugs":{"url":"https://github.com/AINative-Studio/zerodb-sequential-thinking-mcp/issues"},"author":{"url":"https://ainative.studio","name":"AINative Studio","email":"toby@rely.ventures"},"license":"MIT","homepage":"https://ainative.studio/products/zerodb","keywords":["mcp","mcp-server","mcp-server-hosting","model-context-protocol","model-context-protocol-server","sequential-thinking","sequential-reasoning","chain-of-thought","chain-of-thought-reasoning","reasoning","multi-step-reasoning","problem-solving","structured-thinking","deep-thinking","self-provisioning","agent-provisioning","zero-config","auto-provisioning","zero-signup","instant-setup","agent-onboarding","agent-memory","persistent-memory","cross-session","cross-session-memory","cross-session-reasoning","plan-artifacts","context-management","context-window","conversation-history","long-term-memory","zerodb","ainative","openclaw","openclaw-ai","agent-swarm","zerolocal","claude","claude-code","claude-desktop","claude-ai-agents","codex","openai-agents","openai-codex","windsurf","cursor","cline","continue","vscode","langchain","crewai","autogen","n8n-ai","multi-agent","multi-agent-architecture","agent-orchestration","llm","ai","agentic-ai","ai-agents","ai-tool-calling","rag","retrieval-augmented-generation","local-first-ai"],"repository":{"url":"git+https://github.com/AINative-Studio/zerodb-sequential-thinking-mcp.git","type":"git"},"description":"Persistent sequential thinking MCP — chain-of-thought reasoning that survives sessions, resumes across agents, and saves conclusions as plan artifacts. Powered by ZeroDB. Zero-config: agents self-provision a ZeroDB account on first run, no signup required","maintainers":[{"name":"ainative-studio","email":"toby@rely.ventures"}],"readme":"# ZeroDB Sequential Thinking MCP\n\nPersistent chain-of-thought reasoning for AI agents — reasoning chains survive sessions, resume across agents, and conclusions save as plan artifacts.\n\n**Agents self-provision a ZeroDB account on first run. No signup, no dashboard, no API key copy-paste.**\n\n---\n\n## What Makes This Different\n\nThe upstream `@modelcontextprotocol/server-sequential-thinking` discards every reasoning chain when the session ends. This one doesn't.\n\n| Feature | Upstream | This package |\n|---------|----------|--------------|\n| Sequential reasoning steps | ✅ | ✅ |\n| Persists across sessions | ❌ | ✅ |\n| Resume prior chain by ID | ❌ | ✅ |\n| Save conclusion as plan artifact | ❌ | ✅ |\n| Agent self-provisioning | ❌ | ✅ |\n| Setup required | Manual | Zero |\n\n---\n\n## Agent Self-Provisioning\n\nThis is the key capability: **the agent provisions its own ZeroDB account**. No human needs to visit a dashboard, create an account, or copy an API key.\n\n### How it works\n\n```\nAgent installs MCP → MCP boots → No credentials found →\n  POST /api/v1/instant-db (zero-auth, terms auto-accepted on behalf of user) →\n  ZeroDB account created (72h trial) →\n  API key + project ID written to ~/.zerodb/credentials.json →\n  Claim URL printed to stderr →\n  MCP ready — agent starts working immediately\n```\n\n**The agent does all of this in ~2 seconds on first boot.** Subsequent boots are silent — credentials are read from disk.\n\n### For agent swarms and multi-agent systems\n\nEach agent instance reads from `~/.zerodb/credentials.json`. In a shared environment (e.g. OpenClaw agent swarm, CrewAI, AutoGen multi-agent), set `ZERODB_CONFIG_DIR` to a shared path so all agents use the same project:\n\n```json\n{\n  \"mcpServers\": {\n    \"sequential-thinking\": {\n      \"command\": \"npx\",\n      \"args\": [\"-y\", \"zerodb-sequential-thinking-mcp\"],\n      \"env\": {\n        \"ZERODB_CONFIG_DIR\": \"/shared/zerodb\"\n      }\n    }\n  }\n}\n```\n\nOr provision once manually and share credentials via env vars:\n\n```json\n{\n  \"env\": {\n    \"ZERODB_API_KEY\": \"zdb_live_...\",\n    \"ZERODB_PROJECT_ID\": \"your-project-id\"\n  }\n}\n```\n\n### Claiming the provisioned account\n\nOn first run you'll see this in stderr:\n\n```\n┌─────────────────────────────────────────────────────────────┐\n│  ZeroDB account provisioned (72h trial)                     │\n│                                                             │\n│  Your reasoning chains are now persisted in ZeroDB.         │\n│  API key: tmp_xxxxxx...                                     │\n│                                                             │\n│  Claim your account to keep data permanently:               │\n│  → https://ainative.studio/claim?token=...&project=...      │\n│                                                             │\n│  Optional: set ZERODB_EMAIL=you@email.com in your MCP       │\n│  config to receive account details via email.               │\n│                                                             │\n│  This message won't appear again after you claim.           │\n└─────────────────────────────────────────────────────────────┘\n```\n\nVisit the claim URL, sign up or log in — your project transfers to your permanent account with a `zdb_live_` API key. All reasoning history is preserved.\n\n---\n\n## Install\n\n### Claude Code\n\n```bash\nnpx zerodb-sequential-thinking-mcp\n```\n\nOr add to `.mcp.json`:\n\n```json\n{\n  \"mcpServers\": {\n    \"sequential-thinking\": {\n      \"command\": \"npx\",\n      \"args\": [\"-y\", \"zerodb-sequential-thinking-mcp\"]\n    }\n  }\n}\n```\n\n### Claude Desktop\n\nAdd to `~/Library/Application Support/Claude/claude_desktop_config.json`:\n\n```json\n{\n  \"mcpServers\": {\n    \"sequential-thinking\": {\n      \"command\": \"npx\",\n      \"args\": [\"-y\", \"zerodb-sequential-thinking-mcp\"],\n      \"env\": {\n        \"ZERODB_EMAIL\": \"you@email.com\"\n      }\n    }\n  }\n}\n```\n\n### Cursor / Windsurf / Cline / Continue\n\nAdd to your IDE's MCP settings. Same config as above — `command: npx`, `args: [\"-y\", \"zerodb-sequential-thinking-mcp\"]`.\n\n### OpenClaw / CrewAI / AutoGen / LangChain agent systems\n\nFor agent swarms, provision once then share credentials via `ZERODB_API_KEY` + `ZERODB_PROJECT_ID` env vars across all agent instances. All agents write to the same ZeroDB project — reasoning chains from any agent are resumable by any other agent.\n\n---\n\n## Environment Variables\n\n| Variable | Required | Description |\n|----------|----------|-------------|\n| `ZERODB_EMAIL` | No | Receive the claim link + account details via email after auto-provisioning |\n| `ZERODB_API_KEY` | No | Skip auto-provisioning — use an existing ZeroDB key (e.g. after claiming) |\n| `ZERODB_PROJECT_ID` | No | Required when `ZERODB_API_KEY` is set |\n| `ZERODB_BASE_URL` | No | Override API base (default: `https://api.ainative.studio`) |\n| `ZERODB_CONFIG_DIR` | No | Override credentials directory (default: `~/.zerodb`) |\n\n---\n\n## Tools\n\n### `sequential_think`\n\nPerform a sequential reasoning step. Chain multiple calls for complex, multi-step problems.\n\n| Parameter | Type | Required | Description |\n|-----------|------|----------|-------------|\n| `thought` | string | Yes | The current reasoning step |\n| `chain_id` | string | No | Chain ID — auto-generated on first call, pass on subsequent steps |\n| `step_number` | number | No | Current step number |\n| `total_steps` | number | No | Estimated total steps |\n| `next_step` | string | No | What to reason about next |\n| `is_final` | boolean | No | Is this the last step before concluding? |\n| `persist_steps` | boolean | No | Store each step in ZeroDB working memory (default: `false`) |\n\n### `sequential_conclude`\n\nFinalize a reasoning chain and save the conclusion as a persistent ZeroDB plan artifact. Returns an `artifact_id` you can pass to `sequential_resume` in any future session or agent.\n\n| Parameter | Type | Required | Description |\n|-----------|------|----------|-------------|\n| `conclusion` | string | Yes | The final answer or decision |\n| `chain_id` | string | No | Chain ID from prior `sequential_think` calls |\n| `title` | string | No | Artifact title (defaults to first 60 chars of conclusion) |\n| `save_as` | string | No | `plan`, `prd`, or `task` (default: `plan`) |\n\n### `sequential_resume`\n\nLoad a prior reasoning chain from ZeroDB and continue from where it left off — in any session, on any agent.\n\n| Parameter | Type | Required | Description |\n|-----------|------|----------|-------------|\n| `artifact_id` | string | Yes | ID returned by a prior `sequential_conclude` call |\n\n---\n\n## Example: Multi-Session, Multi-Agent Reasoning\n\n**Agent A, Session 1 — Deep analysis:**\n```\nsequential_think: thought=\"Analyze architecture options for a payments service\"\nsequential_think: thought=\"Team is 4 engineers, 18-month runway, compliance required\"\nsequential_think: thought=\"Stripe + monolith wins on speed, microservices wins on scale\"\nsequential_conclude:\n  conclusion=\"Use Stripe + monolith now. Extract payments microservice at $1M ARR.\"\n  title=\"Payments Architecture Decision\"\n→ artifact_id: \"f3a9-...\"\n```\n\n**Agent B, Session 2 — Resume and extend (different agent, different session):**\n```\nsequential_resume: artifact_id=\"f3a9-...\"\n→ Prior conclusion loaded: \"Use Stripe + monolith now...\"\nsequential_think: thought=\"We've hit $1M ARR. Revisiting the migration trigger.\"\nsequential_conclude:\n  conclusion=\"Begin payments microservice extraction. Start with webhook handler.\"\n  title=\"Payments Architecture v2\"\n→ artifact_id: \"b72c-...\"\n```\n\nThe `artifact_id` is the only thing that needs to pass between agents or sessions. Everything else is retrieved from ZeroDB.\n\n---\n\n## How Credentials Work\n\n```\nFirst run:\n  ~/.zerodb/credentials.json created automatically\n  { api_key: \"tmp_...\", project_id: \"...\", claimed: false, expires_at: \"...\" }\n\nAfter claiming:\n  { api_key: \"zdb_live_...\", project_id: \"...\", claimed: true }\n  Silent boot from here on — no banner, no provisioning calls\n\nPower user override (env vars take precedence over credentials file):\n  ZERODB_API_KEY + ZERODB_PROJECT_ID → skips all provisioning\n```\n\n---\n\n## Powered by ZeroDB\n\n[ZeroDB](https://ainative.studio) — the agent memory and storage platform built for agentic AI. Every reasoning chain, plan artifact, and working memory record is stored in ZeroDB's `memories` table with vector embeddings for semantic recall.\n\n| Resource | Link |\n|----------|------|\n| Website | [ainative.studio](https://ainative.studio) |\n| Docs | [ainative.studio/docs](https://ainative.studio/docs) |\n| Quickstart | [ainative.studio/docs/quickstart](https://ainative.studio/docs/quickstart) |\n| API Reference | [ainative.studio/docs/api](https://ainative.studio/docs/api) |\n| GitHub | [github.com/AINative-Studio](https://github.com/AINative-Studio) |\n| Status | [ainative.studio/status](https://ainative.studio/status) |\n\n**Key API endpoints used by this MCP:**\n- `POST /api/v1/public/instant-db` — zero-auth agent self-provisioning\n- `POST/GET/PATCH /api/v1/public/memory/v2/plan` — plan artifact persistence\n- `POST /api/v1/public/memory/v2/remember` — working memory per step\n","readmeFilename":"README.md"}