Wisegit
Give your agent git-aware context so it respects intentional design decisions instead of blindly rewriting sensitive code during AI edits.
Overview
wisegit is a Ship-phase MCP server that surfaces git-derived decision intent to reduce harmful AI edits to intentional code.
What is this MCP server?
- Extracts decision intent from git commit history for agent context
- Helps shield intentional or delicate code from careless AI modifications
- stdio MCP via npm package @sandip124/wisegit
- Version 0.1.1 with GitHub source at Sandip124/wisegit
- Focused git guardrail server, not a full code-review platform
- Server version 0.1.1
- Single npm package @sandip124/wisegit with stdio transport
Community signal: 1 GitHub stars.
What problem does it solve?
Agents rewrite code without knowing prior tradeoffs, so fragile or deliberate implementations get smoothed over in automated edits.
Who is it for?
Solo developers using heavy AI refactors who want git-backed guardrails on sensitive modules.
Skip if: Greenfield repos with almost no commit narrative or teams that do not use git with agents.
What do I get? / Deliverables
With wisegit registered, your agent can align suggestions with recorded intent from git history before you accept changes.
- MCP-accessible git decision-intent signals
- Reduced risk of agent edits against documented choices
- stdio server registered in agent config
Recommended MCP Servers
Journey fit
How it compares
Git-intent MCP guardrail, not a pull-request review bot or static analyzer skill.
Common Questions / FAQ
Who is wisegit for?
Indie builders and small teams who use MCP-enabled agents on real git repos and worry about AI overwriting intentional design.
When should I use wisegit?
Use it in Ship during review—before merging agent-generated diffs or during iterative fixups on legacy areas with rich commit messages.
How do I add wisegit to my agent?
Install @sandip124/wisegit from npm, add a stdio MCP server entry pointing at that package, and restart Claude Code, Cursor, or your MCP host.