Query Hlido from any MCP client
Hlido publishes the trust layer for AI agents — C2PA-verified reviews, Laddoo Scores, tier verdicts. All of it is queryable over Model Context Protocol at a single HTTPS endpoint. No auth, no rate limit abuse, no scraping required.
Endpoint
List available tools
Example: trust_check for an agent
Tools
trust_check
One-call trust summary for any reviewed agent. Pass agent_or_url (slug or full URL). Returns score, tier, blockers, verdict, review URL, embed URL, and C2PA note.
find_trusted
Ranked matches for a free-text need. Pass need, optional min_tier (default STEADY), optional limit (default 10, max 25).
verify_claim
Check whether a specific claim has been tested by Hlido. Returns honest_null when the claim is not specifically addressed in a current review — we don't pretend otherwise.
compare_agents
Side-by-side comparison of up to 5 agents by slug.
submit_agent
Queue a new agent for Hlido review. Pass url and name, plus optional note. Hlido reviews are independent — submission is not purchase.
Registry listings
Hlido is also listed in: smithery.ai, mcp.so, and the modelcontextprotocol/servers registry. Server card: /.well-known/mcp-server-card/server.json.
Disclosure
Hlido reviews are editorial and independent. The Verified Badge is a separate subscription product and does not influence scoring. Sanitized metadata is mirrored to hlido-eu/agent-benchmark under CC-BY-4.0; full evidence and attestations remain on hlido.eu.