—
/100
Unscored
○ Unscored 0⁄0
openclaw-output-vetter-mcp
MCP server for verifying AI agent claims vs reality — single-transcript inline grounding-check that flags when an agent's response states facts not in the input context, when its code silently swallows exceptions and substitutes mock data, or when its multi-turn transcript contains contradictions or unverified completion claims. Sub-second, local, free, no API calls.
Glama
Unscored visibility
— 0/0 applicable dimensions scored
○ Schema Quality
— Protocol
— Reliability
○ Docs & Maintenance
○ Security Hygiene
— Schema Interpretability
Schema Quality
—
42% weight
Protocol Compliance
N/A
Local server
Reliability
N/A
Local server
Docs & Maintenance
—
25% weight
Security Hygiene
—
33% weight
Add this to your README to display your MCP Scoreboard grade:
