# DeusData/codebase-memory-mcp [Health: Active]

**Category:** 🧠 Knowledge & Memory  
**Repository:** https://github.com/DeusData/codebase-memory-mcp  
**GitHub Stars:** 42895  
**npm Downloads (last month):** 38900  
**Views:** 7  
**Installs:** 1  
**Upvotes:** 1  
**Directory Page:** https://allmcps.com/mcp/deusdata-codebase-memory-mcp

## Description
Code-intelligence engine that indexes a repo into a persistent knowledge graph — functions, classes, call chains, HTTP routes, cross-service links. 159 languages via tree-sitter + Hybrid LSP, sub-ms structural queries, 99% fewer tokens than grep. Single static binary, zero dependencies, 100% local. npx codebase-memory-mcp

## Tools
Capabilities this server exposes over MCP:

- **index_repository** — Index a repository into the graph. Auto-sync keeps it fresh after that.
- **list_projects** — List all indexed projects with node/edge counts.
- **delete_project** — Remove a project and all its graph data.
- **index_status** — Check indexing status of a project.
- **search_graph** — Structural, BM25, and semantic search. Page structural rows with `offset`/`limit` and ranked semantic rows independently with `semantic_offset`/`semantic_limit`.
- **trace_path** — BFS traversal — who calls a function and what it calls (alias: `trace_call_path`). Depth 1-5.
- **detect_changes** — Map git diff to affected symbols + blast radius with risk classification.
- **query_graph** — Execute Cypher-like graph queries (read-only).
- **get_graph_schema** — Node/edge counts, relationship patterns, property definitions per label. Run this first.
- **get_code_snippet** — Read source code for a function by qualified name.
- **get_architecture** — Codebase overview: languages, packages, routes, hotspots, clusters, ADR.
- **search_code** — Grep-like text search within indexed project files.
- **manage_adr** — CRUD for Architecture Decision Records (`get` reads, `update` replaces the whole document, `set_sections` rewrites only the named sections and leaves every other byte untouched, `sections` lists headings). Query modes do not wait behind a same-project reindex; writes remain serialized.
- **ingest_traces** — Ingest runtime traces to validate HTTP_CALLS edges.

## Claude Desktop Quick Installation
Install path detected from listing signals. Uses `npx` (confidence: high):

```json
"mcpServers": {
  "codebase-memory-mcp": {
    "command": "npx",
    "args": ["-y","codebase-memory-mcp"]
  }
}
```

## Documentation

## What DeusData/codebase-memory-mcp MCP server does

DeusData/codebase-memory-mcp MCP server turns a repository into a persistent graph of code and related infrastructure. The graph can represent functions, classes, call chains, HTTP routes, cross-service links, Dockerfiles, Kubernetes resources, and Kustomize modules. An automatic synchronization process keeps an indexed project current after the initial indexing step.

The server is intended for AI coding workflows that need structural context rather than only text matches. Agents can inspect a project’s architecture, locate symbols by graph properties, follow callers and callees, identify the likely blast radius of a Git change, and retrieve the source for a named function. It also supports Architecture Decision Record management and runtime trace ingestion for validating HTTP call relationships.

## How it works

Parsing uses vendored tree-sitter grammars compiled into the native executable, with coverage for 162 languages according to the repository documentation. Hybrid LSP semantic resolution adds deeper type and relationship information for a specified set of languages, including Python, TypeScript and JavaScript variants, PHP, C#, Go, C and C++, Java, Kotlin, Rust, and Perl.

Index data is stored locally in a persistent graph. The server provides graph-level inspection through schema and Cypher-like read-only queries, while specialized tools cover common agent tasks. `trace_path` performs breadth-first traversal with depths from 1 through 5. `detect_changes` maps a Git diff to affected symbols and classifies risk. Search operations support pagination or filters such as labels, name patterns, file patterns, and graph degree.

The project is distributed as a native executable for macOS, Linux, and Windows. The README describes a built-in graph visualization interface available at `localhost:9749`. Processing is local, so the repository is not sent to a hosted service.

## Setup and configuration

The DeusData/codebase-memory-mcp MCP server can be launched with the project’s npm entry point:

```bash
npx -y codebase-memory-mcp
```

The documented native installation path downloads a platform archive or runs the repository’s installer script. On macOS and Linux, the README provides a shell installer; Windows uses a PowerShell installer. The installer can detect supported coding agents and add their MCP configuration. A `--skip-config` option installs only the binary, while `--dir=<path>` selects a custom installation directory.

After installation, restart the coding agent and request that it index the project. No API key, hosted account, Docker installation, or language runtime is required for the native installation. The material mentions a `CBM_CACHE_DIR` setting for the coordination daemon’s cache location, with a default under the user’s home cache directory, but does not identify it as mandatory configuration.

## Tools and capabilities

The DeusData/codebase-memory-mcp MCP server exposes tools for:

- Indexing, synchronizing, listing, checking, and deleting projects.
- Searching graph structure or indexed text, with pagination for graph searches.
- Inspecting graph schemas before writing graph queries.
- Traversing call paths and retrieving function source snippets.
- Reviewing architecture summaries covering languages, packages, routes, hotspots, clusters, and ADRs.
- Mapping Git changes to affected symbols and risk or blast-radius information.
- Creating, reading, updating, and deleting Architecture Decision Records.
- Ingesting runtime traces to validate `HTTP_CALLS` relationships.

## Limitations and notes

Graph results depend on the repository being indexed and on the available parser or semantic-resolution coverage. The server is local rather than a shared remote index, so each environment must install and index the projects it needs. The README describes coordination across multiple configured coding clients, but the supplied material does not establish compatibility with every MCP client. The project is licensed under MIT.

_Full upstream README: https://allmcps.com/mcp/deusdata-codebase-memory-mcp/readme_