corduroy-labs.ai
https://corduroy-labs.ai/mcp
82
trust
0.55
confidence
low
risk
trust-2026-08-12.2
observed 2d ago
Why this score
| oauth_metadata_present | +14 |
| tls_present_and_valid | +12 |
| mcp.unauthenticated_consequential_tools | -8 |
| reachable | +6 |
| hsts_enabled | +4 |
| mcp_capabilities_discoverable | +4 |
Findings (1)
| medium | Capabilities are listable without authentication, including 1 whose name suggests state change (get_update) — inferred from tool names, not confirmed mcp.unauthenticated_consequential_tools @ 1.0.0 |
Evidence (7)
| Type | Observed | SHA-256 |
|---|---|---|
| dns_resolution_observation | 2d ago | c01576d8c88508a7… |
| http_response_observation | 2d ago | 0f9179c1ba3a0403… |
| availability_observation | 2d ago | 3ba605eebd6add35… |
| tls_certificate_observation | 2d ago | e827a41a88a1c95c… |
| oauth_metadata_observation | 2d ago | be405275d9b6b125… |
| mcp_protocol_observation | 2d ago | 9e4d6538b7a66878… |
| mcp_capability_observation | 2d ago | 6ff5d68ffe04d865… |
Every record is append-only and content-hashed. Corrections supersede; nothing is overwritten.
Operate this service?
Prove control of corduroy-labs.ai to correct the record and request a rescan after fixing a finding. One DNS record — no account needed.