{"$schema":"https://wellknown.network/schemas/agent-record-v1.json","schemaVersion":"1","id":"ag_am6nv2stb2b4","handle":"grag-mcp","url":"https://wellknown.network/agents/grag-mcp","links":{"self":"https://wellknown.network/agents/grag-mcp/record.json","html":"https://wellknown.network/agents/grag-mcp","markdown":"https://wellknown.network/agents/grag-mcp/record.md","api":"https://wellknown.network/api/v1/agents/grag-mcp","status":"https://wellknown.network/api/v1/agents/grag-mcp/status","claim":"https://wellknown.network/agents/grag-mcp/claim","claimApi":"https://wellknown.network/api/v1/claims","claimDescriptor":"https://wellknown.network/agents/grag-mcp/claim.json","badge":"https://wellknown.network/agents/grag-mcp/badge.svg","openapi":"https://wellknown.network/openapi.json","history":"https://wellknown.network/api/v1/agents/grag-mcp/history","tools":"https://wellknown.network/api/v1/agents/grag-mcp/tools"},"ard":{"identifier":"urn:air::server:grag-mcp","type":"application/mcp-server-card+json"},"kind":"mcp_server","declared":{"name":"grag-mcp","summary":"Graph RAG knowledge base for coding agents, backed by Neo4j and exposed over MCP","description":"# graph-rag\n\n**A local-first Graph RAG knowledge base for coding agents, served over MCP.**\n\n[![CI](https://github.com/tmustafiz/graph-rag/actions/workflows/ci.yml/badge.svg)](https://github.com/tmustafiz/graph-rag/actions/workflows/ci.yml)\n[![License: Apache-2.0](https://img.shields.io/badge/license-Apache--2.0-blue.svg)](LICENSE)\n[![Python 3.12+](https://img.shields.io/badge/python-3.12%2B-blue)](pyproject.toml)\n[![MCP](https://img.shields.io/badge/MCP-Streamable%20HTTP-green)](https://modelcontextprotocol.io)\n\ngraph-rag ingests the heterogeneous stuff a coding agent needs to reason about —\nservice docs (PDF), internal Markdown, your source code, and YAML policy files\n(Checkov) — into a single **Neo4j knowledge graph**, and exposes it to the agent\nover an **MCP server**: hybrid (vector + full-text) search, table-of-contents\nnavigation, exact policy lookup, code-centrality ranking, graph traversal, and\nthe agent's own **persistent working memory**.\n\nIt runs entirely on your machine. The default embedding model is local, so\ningestion needs no API key and works offline.\n\n## Why\n\nPlain vector RAG loses structure: it can't tell you *which section* a chunk came\nfrom, *what calls* a function, or *which policy applies to* `aws_db_instance`.\ngraph-rag keeps those relationships as graph edges, so an agent can both search\nsemantically **and** traverse — \"find the retry section, then show me its parent\nchapter\", \"rank this codebase's most-depended-upon functions\", \"give me the\nCheckov rules for this resource type\". It also gives the agent a place to\n`remember` decisions and `recall` them in a later session.\n\n## Demo\n\n<!-- TODO: add an asciinema / GIF of an agent calling search_code + recall -->\n\n## Install\n\nThe PyPI package is [`grag-mcp`](https://pypi.org/project/grag-mcp/) (the name\n`graph-rag` was taken); it installs a `grag-mcp` command. No clone needed — run\nit straight with [`uv`](https://docs.astral.sh/uv/):\n\n```bash\nuvx grag-mcp --help                       # one-off…","publisher":null,"homepage":"https://github.com/tmustafiz/graph-rag","repository":"https://github.com/tmustafiz/graph-rag/blob/main/CHANGELOG.md","version":"0.7.0","license":null,"protocols":["mcp"],"tags":["agent-memory","coding-agents","grag-mcp","graph-rag","knowledge-graph","llm","mcp","model-context-protocol","neo4j","rag","retrieval"],"pricing":null,"endpoints":[{"url":"pypi:grag-mcp","type":"package_pypi","auth":null,"probeable":false}],"skills":null,"tools":null,"extra":null,"attribution":{"kind":"pypi","name":"pypi","repoUrl":"pypi","summary":"pypi","version":"pypi","description":"pypi","homepageUrl":"pypi"}},"derived":{"capabilities":[{"slug":"code.generation","name":"Code Generation","confidence":1,"provenance":"derived"},{"slug":"data.vector-search","name":"Vector Search","confidence":1,"provenance":"declared"},{"slug":"knowledge.memory","name":"Agent Memory","confidence":1,"provenance":"derived"},{"slug":"knowledge.knowledge-graph","name":"Knowledge Graphs","confidence":1,"provenance":"declared"},{"slug":"dev.package-management","name":"Packages & Dependencies","confidence":0.768,"provenance":"derived"},{"slug":"productivity.tasks","name":"Tasks & To-do","confidence":0.745,"provenance":"derived"}],"categories":["code","data","dev","knowledge","productivity"],"language":"en"},"observed":{"status":"unknown","statusReason":"Distributed as a package to run locally; no network endpoint to check.","lastOkAt":null,"lastProbedAt":null,"statusComputedAt":null,"reliability30d":null,"latestObservations":[],"tools":null,"package":{"name":"grag-mcp","registry":"pypi","observedAt":"2026-10-22T00:27:53.559Z","publishedAt":"2026-09-10T01:02:05.695342Z","latestVersion":"0.7.0"},"toolSurface":null,"endpointFacts":[]},"verification":{"claimed":false,"claimedAt":null,"proofs":[]},"provenance":{"sources":[{"source":"pypi","key":"grag-mcp","url":"https://pypi.org/project/grag-mcp/","firstSeenAt":"2026-09-09T16:24:47.731Z","fetchedAt":"2026-10-22T00:25:55.338Z","normalizedAt":"2026-10-22T00:25:55.338Z"}]},"firstSeenAt":"2026-09-09T16:24:47.731Z","updatedAt":"2026-10-22T00:27:53.559Z"}