com.obeliskgate/trust-tools

Measured conclusion: Deep-tested. L2 outcome: tools_ok.

Website security ratings, token verification, and tamper-evident ledger heads from Obelisk Gate.

Add to your agent

Copy this into your MCP client config:

{
  "mcpServers": {
    "trust-tools": {
      "url": "https://obeliskgate.com/mcp",
      "transport": "streamable-http"
    }
  }
}

Measured tools (9)

• scan_trust

  Run Obelisk's public trust scan on an https URL — TLS and security-header posture, scored as an Obelisk Rating. Read-only, SSRF-guarded.

  Parameters:

  • url (required) — string — The https URL to scan, e.g. https://example.com

  Hints: read-only not declared · destructive not declared · idempotent not declared · open-world not declared

• verify_token

  Verify that a JSON Web Token was minted by this Obelisk Gate (ES256, correct issuer) and report its type, subject, assurance, and principal. Never returns secrets.

  Parameters:

  • token (required) — string — A compact-serialized JWT issued by Obelisk (id_token or access token).

  Hints: read-only not declared · destructive not declared · idempotent not declared · open-world not declared

• get_org_rating

  Read the public Obelisk Rating (0-100 score, trust band, and trend) for a registered organization by its slug.

  Parameters:

  • org (required) — string — The organization slug, e.g. acme.

  Hints: read-only not declared · destructive not declared · idempotent not declared · open-world not declared

• verify_agent_run_proof

  Verify an Obelisk run proof. Pass run_proof_token, the token Obelisk signs at POST /api/agent-proof/run: the tool checks Obelisk's signature against its published keys, that the named agent is active now with the same key that answered the run challenge, and which delegation hops matched the owner's registry when the token was signed. A bare run_proof envelope is only checked for internal consistency and is never reported as verified, because anyone can build one. Reports the envelope's commitment scheme: a legacy v1 envelope, whose hashes are unsalted, carries the warning unsalted-v1. Does not reveal or infer raw task data.

  Parameters:

  • run_proof — object — An obelisk-agent-run-proof-v2 envelope (a legacy v1 envelope is accepted and flagged unsalted-v1). Alone it is checked for consistency only; next to run_proof_token it must equal the signed copy.
  • run_proof_token — string — The run_proof_token returned by POST /api/agent-proof/run. The only input that can verify.
  • expected_agent_id — string — Optional expected agt_ subject.
  • expected_audience — string — Optional: require the run token's audience to equal this value.

  Hints: read-only not declared · destructive not declared · idempotent not declared · open-world not declared

• get_agent_proof

  Read an Obelisk agent's public proof vector by opaque proof id. Returns independent claims with freshness and scope; never a scalar trust score or a claim of non-humanness.

  Parameters:

  • proof_id (required) — string — The opaque id from an /agent/<proof_id> link.

  Hints: read-only not declared · destructive not declared · idempotent not declared · open-world not declared

• transparency_head

  Read the current signed transparency head of Obelisk's tamper-evident receipt ledger. Save headAnchorHash with anchorCount and compare later: the count must never go down, and the head at a saved count must not change. No consistency proofs are published, so this cannot show a later head extends a saved one. Same data as /.well-known/obelisk-transparency.json.

  Hints: read-only not declared · destructive not declared · idempotent not declared · open-world not declared

• gate_status

  Read the Gate's public liveness facts: issuer, served code revision, OIDC availability, and supported protocols. No posture internals.

  Hints: read-only not declared · destructive not declared · idempotent not declared · open-world not declared

• pin_tool_manifest

  Pin an MCP tool manifest before an agent trusts it. Returns a SHA-256 digest per tool and for the whole catalogue (canonical JSON, every field counts), hidden-text findings (Unicode tag, bidi and zero-width characters) and an injection signal per tool. Pass the digests you stored at approval as expected to detect a changed, added or removed tool. Up to 64 tools in a request body of at most 256 KB. Read-only and stateless; the injection score is one signal, not a guarantee.

  Parameters:

  • tools (required) — array — The tools array exactly as the server's tools/list returned it.
  • expected — Optional: the catalogueDigest string from an earlier pin, or an object of { toolName: digest } (optionally with catalogue).

  Hints: read-only not declared · destructive not declared · idempotent not declared · open-world not declared

• explain_rating

  Explain what an Obelisk Rating number means: its trust band, how the public scan scores, and what typically moves a score. Deterministic by default; pass narrate:true for an additional one-line model-written narrative (cached, public-scope, never required). Pure function of the number — no data is read.

  Parameters:

  • rating (required) — number — A 0-100 Obelisk Rating, e.g. from scan_trust or get_org_rating.
  • narrate — boolean — S283 — also return `narrative`, a one-sentence model-written reading of the band (deterministic fields are always present).

  Hints: read-only not declared · destructive not declared · idempotent not declared · open-world not declared

Publisher website

Measured by MCPYP probing. Untested fields are omitted, never estimated. Listing is always free; MCPYP never proxies calls or stores user tokens.