This one watches your AI assistant conversations for scope creep against project contracts. The idea is to catch when requests drift beyond the agreed statement of work and flag violations in real time. It also drafts change order responses when out of scope work comes up, which could save some awkward client conversations. The GitHub repo is pretty bare bones with no readme or implementation details visible, so it's hard to say what contract formats it supports or how the detection logic actually works. If you're freelancing or running an agency where scope boundaries matter, the concept makes sense. You'd want to dig into the code before relying on it for billable work though.
claude mcp add --transport stdio io.github.dgtalquantumleap-ai-ebenova-scope-guard -- npx -y @ebenova/scope-guard-mcp