{"_id":"@agentica/pg-selector","_rev":"4-efcdcf9be3117dcee9f46b25b57a6c15","name":"@agentica/pg-selector","dist-tags":{"latest":"0.10.3"},"versions":{"0.10.3":{"name":"@agentica/pg-selector","version":"0.10.3","keywords":["openai","chatgpt","anthropic","claude","ai","chatbot","nestia","swagger","openapi"],"author":{"name":"Wrtn Technologies"},"license":"MIT","_id":"@agentica/pg-selector@0.10.3","maintainers":[{"name":"samchon","email":"samchon.github@gmail.com"}],"homepage":"https://wrtnlabs.io","bugs":{"url":"https://github.com/wrtnlabs/agent/issues"},"dist":{"shasum":"a2485a779d7f62abc7f59d688e8a956971daf137","tarball":"https://registry.npmjs.org/@agentica/pg-selector/-/pg-selector-0.10.3.tgz","fileCount":16,"integrity":"sha512-Tw2Nl8yp+gaQmErkLUOb3X7YuR7i5NRBa2XY0aMw4n+qCa927DjpjBEx7eTwAHNspuM406HuToffnRicRd1rIQ==","signatures":[{"sig":"MEYCIQDSOncGhtN8Abh35iFZOHI/dO0ZsFMUIK3iOkGSn0k9yAIhAL3RS3x+4eacnM4SOnPVgAWStdJvJTgp5reK8SnYGECy","keyid":"SHA256:DhQ8wR5APBvFHLF/+Tc+AYvPOdTpcIDqOhxsBHRwC7U"}],"unpackedSize":22176},"main":"lib/index.js","module":"lib/index.mjs","gitHead":"3670a762706390b7c1c936263fe420f41d5a16ff","scripts":{"test":"vitest","build":"tsc && rollup -c","eslint":"eslint ./**/*.ts"},"typings":"lib/index.d.ts","_npmUser":{"name":"samchon","email":"samchon.github@gmail.com"},"repository":{"url":"git+https://github.com/wrtnlabs/agent.git","type":"git"},"_npmVersion":"10.8.2","description":"Super A.I. Chatbot agent by Swagger Document","directories":{},"_nodeVersion":"20.18.3","dependencies":{"@wrtnlabs/connector-hive-api":"^1.4.1"},"_hasShrinkwrap":false,"devDependencies":{"eslint":"^9.17.0","rimraf":"^6.0.1","rollup":"^4.29.1","prettier":"^3.5.0","@eslint/js":"^9.17.0","typescript":"~5.7.3","@types/node":"^22.10.5","@agentica/core":"^0.10.3","@samchon/openapi":"^3","typescript-eslint":"^8.18.2","@rollup/plugin-terser":"^0.4.4","@rollup/plugin-typescript":"^12.1.1","@trivago/prettier-plugin-sort-imports":"^5.2.2"},"peerDependencies":{"@agentica/core":"workspace:^"},"_npmOperationalInternal":{"tmp":"tmp/pg-selector_0.10.3_1741161076003_0.3233347256040684","host":"s3://npm-registry-packages-npm-production"}}},"time":{"created":"2025-03-05T07:51:15.941Z","modified":"2026-04-03T09:10:34.760Z","0.10.3":"2025-03-05T07:51:16.167Z"},"bugs":{"url":"https://github.com/wrtnlabs/agent/issues"},"author":{"name":"Wrtn Technologies"},"license":"MIT","homepage":"https://wrtnlabs.io","keywords":["openai","chatgpt","anthropic","claude","ai","chatbot","nestia","swagger","openapi"],"repository":{"url":"git+https://github.com/wrtnlabs/agent.git","type":"git"},"description":"Super A.I. Chatbot agent by Swagger Document","maintainers":[{"email":"samchon.github@gmail.com","name":"samchon"}],"readme":"# pg-selector\n\n[![GitHub license](https://img.shields.io/badge/license-MIT-blue.svg)](https://github.com/wrtnlabs/pg-selector/blob/master/LICENSE)\n[![npm version](https://img.shields.io/npm/v/pg-selector.svg)](https://www.npmjs.com/package/pg-selector)\n\nA library that significantly accelerates AI function selection through vector embeddings.\n\n## Overview\n\n`@agentica/pg-selector` drastically improves function selection speed compared to traditional LLM-based methods. By leveraging vector embeddings and semantic similarity, it can identify the most appropriate functions for a given context multiple times faster than conventional approaches.\n\n```typescript\nimport { Agentica } from \"@agentica/core\";\nimport { AgenticaPgVectorSelector } from \"@agentica/pg-selector\";\n\nimport typia from \"typia\";\n\n\n// Initialize with connector-hive server\nconst selectorExecute = AgenticaPgVectorSelector.boot<\"chatgpt\">(\n  'https://your-connector-hive-server.com'\n);\n\n\nconst agent = new Agentica({\n  model: \"chatgpt\",\n  vendor: {\n    model: \"gpt-4o-mini\",\n    api: new OpenAI({\n      apiKey: process.env.CHATGPT_API_KEY,\n    }),\n  },\n  controllers: [\n    await fetch(\n      \"https://shopping-be.wrtn.ai/editor/swagger.json\",\n    ).then(r => r.json()),\n    typia.llm.application<ShoppingCounselor>(),\n    typia.llm.application<ShoppingPolicy>(),\n    typia.llm.application<ShoppingSearchRag>(),\n  ],\n  config: {\n    executor: {\n      select: selectorExecute,\n    }\n  }\n});\nawait agent.conversate(\"I wanna buy MacBook Pro\");\n```\n\n## How to Use\n\n### Setup\n\n```bash\nnpm install @agentica/core @agentica/pg-selector typia\nnpx typia setup\n```\n\nTo use pg-selector, you need:\n\n1. A running [connector-hive](https://github.com/wrtnlabs/connector-hive) server\n2. `PostgreSQL` database connected to the `connector-hive` server\n3. pgvector extension installed in `PostgreSQL`\n\n### Initialization\n\nFirst, initialize the library with your connector-hive server:\n\n```typescript\nimport { AgenticaPgVectorSelector } from 'pg-selector';\n\nconst selectorExecute = AgenticaPgVectorSelector.boot<YourSchemaModel>(\n  'https://your-connector-hive-server.com'\n);\n```\n\n### Just apply Selector and Start conversation\n\nSelect the most appropriate functions for a given context:\n\n```typescript\nconst agent = new Agentica({\n  model: \"chatgpt\",\n  vendor: {\n    model: \"gpt-4o-mini\",\n    api: new OpenAI({\n      apiKey: process.env.CHATGPT_API_KEY,\n    }),\n  },\n  controllers: [\n    await fetch(\n      \"https://shopping-be.wrtn.ai/editor/swagger.json\",\n    ).then(r => r.json()),\n    typia.llm.application<ShoppingCounselor>(),\n    typia.llm.application<ShoppingPolicy>(),\n    typia.llm.application<ShoppingSearchRag>(),\n  ],\n  config: {\n    executor: {\n      select: selectorExecute,\n    }\n  }\n});\nawait agent.conversate(\"I wanna buy MacBook Pro\");\n```","readmeFilename":"README.md"}