kerbelp/metatron
by Various
Metatron is a self-hosted system that captures a codebase's real implementation decisions — preferred patterns, rejected approaches, edge cases, internal conventions — as structure
MCP
kerbelp/metatron
Added 7 June 2026
Overview
Metatron is a self-hosted system that captures a codebase's real implementation decisions as structured priors. It serves these priors to coding agents over MCP (Model Context Protocol). The goal is to make an agent write code like a senior engineer who already knows the codebase.
Best for
Best for
Teams with established codebases who want to encode institutional knowledge for AI coding agents
Use cases
- Capturing preferred patterns and conventions from a codebase
- Providing coding agents with context on past design decisions
- Documenting rejected approaches and edge cases for future reference
Notes
Metatron is a self-hosted system that captures a codebase’s real implementation decisions as structured priors. It serves these priors to coding agents over MCP (Model Context Protocol). The goal is to make an agent write code like a senior engineer who already knows the codebase.
13 stars on GitHub. Last updated 2026-06-07. Licensed MIT.
Use cases
- Capturing preferred patterns and conventions from a codebase
- Providing coding agents with context on past design decisions
- Documenting rejected approaches and edge cases for future reference
Pros
- Self-hosted ensures codebase data stays private and under your control
- Captures nuanced decisions including edge cases and rejected approaches
- Integrates with MCP to give agents context-aware guidance
Cons
- Very early stage project with limited adoption and documentation
- Requires self-hosting and manual setup of prior capture
- The tool itself is written in Python, which may limit extensibility for non-Python environments
Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.
Pros
- Self-hosted ensures codebase data stays private and under your control
- Captures nuanced decisions including edge cases and rejected approaches
- Integrates with MCP to give agents context-aware guidance
Cons
- Very early stage project with limited adoption and documentation
- Requires self-hosting and manual setup of prior capture
- The tool itself is written in Python, which may limit extensibility for non-Python environments
Pairs with
Other entries in the index that connect to this one. Click through to see the chain.
Aider
Paul Gauthier
Terminal-first AI pair programmer. Edits files in your repo, commits with sensible messages, runs your tests.
Claude Code
Anthropic
Anthropic's terminal-native coding agent. Reads your repo, edits files, runs tests, ships PRs.
Cline
Cline
Open-source autonomous coding agent that lives inside VS Code. BYO model key, watch it work.