Theme
AI Resources
Codebase Memory MCP
Codebase Memory MCP is a local code-intelligence server that turns a repository into a persistent knowledge graph for AI coding agents.
It parses code structure, stores symbols and relationships in SQLite, keeps the graph updated as files change, and exposes search, call-path, dependency, and impact questions through MCP tools. Use this as a first read, not a recommendation. Open the original project before trusting details like terms, limits, privacy, cost, setup, or safety.
What it is
A reusable structural map of the code
The server indexes functions, classes, imports, calls, routes, packages, and other relationships so an agent can query a saved code map instead of rebuilding the same understanding through repeated file searches.
Why it stands out
Broad parsing in a local native build
The project combines broad tree-sitter language coverage with deeper LSP-assisted type resolution for selected languages, incremental updates, cross-repository links, graph queries, and installers for several coding-agent clients.
Availability
Cross-platform releases and package paths
The MIT-licensed project publishes binaries for macOS, Linux, and Windows, plus documented package-manager and client setup paths. Its core indexing and MCP handling run locally.
Why it matters
What makes it useful
Large repositories make coding agents spend time and context rediscovering where code lives and how it connects. A persistent structural graph can answer recurring questions about callers, dependencies, routes, and likely change impact before the agent opens every related file again.
What to know
Where it fits
Use it as a context layer beside an MCP-capable coding agent when the work repeatedly crosses many files, packages, or repositories. It does not replace reading the relevant implementation or running the project tests; the graph reflects what its parsers and language integrations can extract.
Notable points
What stands out
A linked project preprint reports substantially lower token and tool use than file-by-file exploration across its benchmark, with lower overall answer quality in the same comparison. Treat those results as a reason to test the approach on your own repositories, not as a guaranteed saving.
Before using
What to review
The server reads source files across the indexed codebase, writes its graph locally, and can change supported agent configuration files during installation.
Review the installer, the selected project scope, ignored paths, generated graph files, and the permissions of the coding agent that can query them.
Decide whether generated graph files should stay local-only or travel with the repository, especially when the codebase is private.
The project is young and moving quickly. Check current releases and open issues for the operating system, language, repository layout, and MCP client you plan to use.
Reader fit
Who may find it relevant
Developers using MCP-capable coding agents across large or repeatedly visited repositories.
Teams comparing persistent SQLite graphs with readable file-based maps, vector search, or other code-intelligence layers.
Builders who want structural queries and incremental codebase context without sending ordinary indexing work to a hosted service.
Less relevant for small one-off projects or readers looking for a complete coding agent rather than a context server it can call.
Editorial note
Why LifeHubber lists it
LifeHubber lists Codebase Memory MCP because it makes repeated repository understanding a persistent local layer. Readers can compare whether a queryable graph saves enough exploration in their own codebase to justify another background tool, index, and permission boundary.
Source links
Source materials
Reader note
Before relying on this entry
LifeHubber lists entries to help readers inspect AI projects, not to endorse them or prove they are safe, suitable, accurate, maintained, or right for a specific use. We do not verify every entry in depth. Before relying on anything listed, review the original materials, terms, privacy practices, limits, and risks that matter for your situation.
What to explore next
Compare how the repository map is built and kept.
A persistent graph is one way to stop an agent rediscovering the same code structure. Compare it with a readable file-based map or another local SQLite-backed index before choosing the setup that fits the repository.
More in AI Agents
Keep browsing this category
Explore more AI agent projects.
Agent-Reach
Panniantong/Agent-Reach
A CLI routing layer that helps command-capable agents reach web, social, repository, video, RSS, and search sources through ordered backends. For multi-backend channels, its doctor command reports the active backend, while default setup only checks the machine unless system changes are explicitly authorized.
AIPOCH Medical Research Skills
aipoch/medical-research-skills
A curated library of medical research agent skills designed to support evidence review, protocol design, data analysis, and academic writing workflows.
Claude Code Game Studios
Donchitos/Claude-Code-Game-Studios
A multi-agent game-development studio system for Claude Code, organized around specialized agents, workflow skills, hooks, rules, and templates.
