masondelan/selvedge
by Various
Long-term memory for AI-coded codebases. A git blame for AI agents — but for the why. MCP server that captures the agent's reasoning live, in context, as each change is made. Local
MCP
masondelan/selvedge
Added 1 June 2026
Overview
Selvedge is an MCP server that records an AI agent's reasoning as it makes changes to a codebase, storing the rationale in a local SQLite database. It functions like a git blame for AI edits, capturing the why behind each modification without external dependencies.
Best for
Best for
Developers using AI coding agents who need to track and review the rationale behind automated changes.
Use cases
- Audit why an AI agent made a specific code change
- Trace the reasoning behind past AI-driven modifications
- Maintain a live log of agent decisions during development
How to use
Install
pip install "selvedge[semantic]" && selvedge index Tools exposed
log_changeprior_attemptsstale_decisions
Tested with
Claude Code, Cursor, Windsurf, ChatGPT
Notes
Selvedge is an MCP server that records an AI agent’s reasoning as it makes changes to a codebase, storing the rationale in a local SQLite database. It functions like a git blame for AI edits, capturing the why behind each modification without external dependencies.
8 stars on GitHub. Last updated 2026-05-26. Licensed MIT.
Use cases
- Audit why an AI agent made a specific code change
- Trace the reasoning behind past AI-driven modifications
- Maintain a live log of agent decisions during development
Pros
- Zero external dependencies, runs locally with SQLite
- Captures reasoning in real time as changes are made
- Provides a clear audit trail for AI-generated code
Cons
- Limited to MCP-compatible AI agents and workflows
- Requires manual setup to integrate with existing projects
- No built-in support for non-AI or manual code changes
Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.
Pros
- Zero external dependencies, runs locally with SQLite
- Captures reasoning in real time as changes are made
- Provides a clear audit trail for AI-generated code
Cons
- Limited to MCP-compatible AI agents and workflows
- Requires manual setup to integrate with existing projects
- No built-in support for non-AI or manual code changes
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.
Cline
Cline
Open-source autonomous coding agent that lives inside VS Code. BYO model key, watch it work.
Claude Code
Anthropic
Anthropic's terminal-native coding agent. Reads your repo, edits files, runs tests, ships PRs.
Get the free Developer’s Field Guide
A 27-page field guide to the AI coding workflow with Claude. Claude Code, MCP servers, the prompt patterns that work, and what to delegate. Free.
Enter your work email. We send it straight over, plus a few short notes worth knowing. Unsubscribe any time.