"author:preprint-labs" matching MCP servers:

Audits coding agents' claimed actions against the actual Git working tree, blocking completion if any claimed changes are unverified.

Enables AI coding assistants to enforce repository rules and block unauthorized file edits or commands, even after context compaction, by running as a local zero-config MCP server.

Serves a central playbook hub that provides situational guidance to LLM agents, lets clients list available playbooks, fetch procedural guidance, and apply graph mutations for persistent self-evolution.

Enables automatic auditing and pruning of uncommitted git diffs to enforce surgical-only AI code changes, removing bloat, drive-by refactors, and cosmetic formatter churn before commit.