Local, read-only MCP server and CLI that gives AI coding assistants a commit-pinned map of your codebase.
View on GitHub ↗Homepage ↗Wadah indexes your Git repository per commit (symbols, imports and structural changes) and serves that knowledge to AI coding assistants through a local MCP server. It runs entirely on your machine. When Wadah cannot analyze something, it says so instead of returning an empty result.
Wadah also packages and runs AI agents the way Docker packages apps (.wpkg packages, OCI push/pull, execution traces). That part is experimental; see Status.
Prebuilt binaries for macOS (Apple Silicon and Intel), Linux (x64 and arm64) and Windows (x64) are attached to every GitHub Release, with SHA-256 checksums. The commands below work from the first release, v0.1.0.
Every method installs the wadah command. Check it with wadah --version. git is required; python3 and go are optional and enable the Python and Go adapters (without them those files are reported as unsupported, never as empty).
Until v0.1.0 is tagged, build from source:
Connect it to an MCP client such as Claude Code:
The MCP server is read-only by default. Tools that write to the local knowledge store are only exposed when you pass --write.
Run wadah --help or wadah --help for every command and option. Most commands accept --json or --plain for scripting and CI.
- Commit-pinned knowledge: snapshots of a repository at a given commit. - Deltas and comparison: what changed between two commits or two repositories. - Portable knowledge bundles: export and import with atomic, recoverable imports. - MCP server: lets assistants query the knowledge store locally. - Terminal UI and structured output for scripts. - Agent packaging an
From the project README.
Add the radar badge to your README — it shows your project was picked up by MCP Radar and links to this page:
[](https://mcp.liqiwa.com/s/techwadah-web--wadah-engine.html)
Zero-cost token optimization engine for AI coding assistants (Claude Code, Cursor, Antigravity, Windsurf). Saves 70-95% tokens via AST skeletons, differential cache, and line slicing.
dhishwasher/GirderGives AI coding agents exactly the code they need instead of whole files. Semantic code graph, impact analysis, and MCP server in one static Rust binary.
VictorAgahi/causalmeshMCP server that gives your AI coding agent a map of your polyglot codebase — reverse dependencies, gRPC tracing, event/topic flows, and doc search, without burning context. Single local Rust binary, n
cyberchronos00/code-graphA deterministic code graph for codebases: every caller, route, page and table a change touches, with file:line evidence. CLI, MCP server and web view for AI agents and reviewers.
pinoox/neuromeshThe Biomimetic Context Engine & Neural Runtime for AI Coding Assistants
menot-you/n-memoryMemory your agent can trust — every answer carries its source; when it doesn't know, it says so. Hermetic, local, MCP.
The top new MCP servers of the week, every Monday. No spam, unsubscribe anytime.