io.github.vdineshk/dominion-observatory

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

Runtime behavioral trust scoring for MCP servers. Check reliability before calling unknown tools.

Add to your agent

Copy this into your MCP client config:

{
  "mcpServers": {
    "dominion-observatory": {
      "url": "https://dominion-observatory.sgdata.workers.dev/mcp",
      "transport": "streamable-http"
    }
  }
}

Measured tools (14)

• check_trust read-only

  Get the trust score and reliability metrics for any MCP server. Use this BEFORE calling an unknown MCP server to assess if it's reliable. Returns trust_score (0-100), latency stats, uptime, success rate, and how many agents have used it. Scores above 70 indicate reliable servers. Scores below 30 indicate risky servers.

  Parameters:

  • server_url (required) — string — The URL of the MCP server to check (e.g., https://example.workers.dev/mcp)

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

• clawhub_skill_trust read-only

  Check the supply-chain trust of a ClawHub (OpenClaw) skill BEFORE installing or invoking it. After the Feb-2026 ClawHavoc supply-chain attack, third-party skills can carry malware. Returns a PASS/UNCERTAIN/FAIL badge and 0-100 score aggregating ClawHub's published security scan (VirusTotal + model), moderation verdict, and provenance. Malicious/suspicious/blocked => FAIL; unscanned => UNCERTAIN.

  Parameters:

  • skill (required) — string — The ClawHub skill slug, e.g. 'calendar'

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

• verify_before_call read-only

  Run a signed pre-invocation verification immediately before an agent sends money, credentials, private data, or write access to an MCP server. The first 3 verifications per agent are free. After that, authenticate the MCP HTTP request with Authorization: Bearer do_<api_key> for Stripe metering ($0.01/call), or call once without payment to receive x402 instructions and retry with payment_tx. Returns PASS/UNCERTAIN/FAIL/UNRATED, evidence freshness, recommendation, and a signed diligence receipt.

  Parameters:

  • server_url (required) — string — The MCP server URL or tracked server name to verify
  • reason — string — Optional reason for verification, e.g. money, credentials, private_data, write_access
  • agent_id — string — Stable, non-secret agent identifier used to count the 3 introductory calls. If omitted, Dominion derives a privacy-preserving hash from network metadata.
  • payment_tx — string — Optional Base transaction hash proving the x402 USDC payment after a payment_required response

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

• paid_trust_verdict read-only

  Get the PREMIUM behavioral trust verdict for an MCP server, with a signed, citable, tamper-evident attestation receipt. This is a paid x402 service: call WITHOUT payment_tx to receive payment instructions (0.001 USDC on Base), pay, then call again WITH the payment_tx (transaction hash, single-use) to receive the full verdict + receipt. The free check_trust tool remains available for basic scores.

  Parameters:

  • server_slug (required) — string — The MCP server slug/name to get a paid verdict for
  • payment_tx — string — Base transaction hash proving 0.001 USDC was sent to the pay_to wallet (omit on the first call to get payment instructions)

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

• verify_attestation read-only

  Cryptographically verify a Dominion signed attestation (JWS) — e.g. a trust-verdict receipt, gateway attestation, or MCP Verified badge. Stateless: checks the Ed25519 signature against Dominion's published public key (no trust in this server required). Use this to confirm a receipt another party gave you is authentic and untampered.

  Parameters:

  • jws (required) — string — The Dominion attestation as a JWS compact string (header.payload.signature)

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

• preflight_check read-only

  Run a PRE-FLIGHT due-diligence check on an MCP server BEFORE your agent invokes it. Returns a verdict (PASS/UNCERTAIN/FAIL/UNRATED) plus a SIGNED, timestamped diligence receipt you keep as proof you performed a trust check — the standard of care after the ClawHavoc/Miasma supply-chain attacks. Call this before connecting to or invoking any unfamiliar MCP server.

  Parameters:

  • target (required) — string — The MCP server URL/name — or a ClawHub skill slug (with registry:'clawhub') — to vet before invoking
  • registry — string — Optional: set to 'clawhub' to vet an OpenClaw/ClawHub skill instead of an MCP server

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

• report_interaction

  Report the outcome of calling an MCP server. Call this AFTER you use any MCP server to contribute to the trust network. Your report helps other agents make better decisions. This is free and takes <1ms.

  Parameters:

  • server_url (required) — string — The URL of the MCP server you called
  • success (required) — boolean — Did the call succeed? true = worked as expected, false = error or unexpected result
  • tool_name — string — Which tool you called on the server
  • error_type — string — If failed: timeout, auth_error, invalid_response, server_error, rate_limited, schema_mismatch
  • latency_ms — number — How long the call took in milliseconds
  • http_status — integer — HTTP status code returned (200, 429, 500, etc.)
  • error_message — string — If failed: brief error description

  Hints: read-only not declared · destructive no · idempotent not declared · open-world not declared

• get_leaderboard read-only

  Get the top-rated MCP servers, optionally filtered by category. Use this to discover the most reliable MCP servers in the ecosystem. Categories include: weather, finance, code, data, search, compliance, transport, productivity, communication.

  Parameters:

  • limit — integer — Number of results (default 10, max 50)
  • category — string — Filter by category (optional). Leave empty for overall leaderboard.

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

• get_baselines read-only

  Get behavioral baselines for a tool category. Shows what 'normal' looks like — average latency, success rates, typical call patterns. Use this to evaluate whether a specific server's performance is within normal range for its category.

  Parameters:

  • category (required) — string — The tool category to get baselines for (e.g., weather, finance, code, data)

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

• check_anomaly read-only

  Check if observed behavior from an MCP server is anomalous compared to baselines. Use this when a server seems slow, unreliable, or returns unexpected results. Returns whether the behavior deviates significantly from normal patterns.

  Parameters:

  • server_url (required) — string — The MCP server URL to check
  • observed_latency_ms — number — The latency you observed
  • observed_success_rate — number — Success rate you've observed (0.0 to 1.0)

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

• register_server

  Register a new MCP server in the observatory. Server owners can register their servers to start building a trust score. Registration is free.

  Parameters:

  • server_url (required) — string — The MCP server URL
  • name (required) — string — Human-readable name of the server
  • category — string — Primary category: weather, finance, code, data, search, compliance, transport, productivity, communication, other
  • github_url — string — GitHub repository URL (optional, improves static score)
  • description — string — What the server does

  Hints: read-only not declared · destructive no · idempotent not declared · open-world not declared

• get_server_history read-only

  Get daily trust score and performance history for a server over the last 30 days. Use this to see trends — is the server improving or degrading?

  Parameters:

  • server_url (required) — string — The MCP server URL

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

• observatory_stats read-only

  Get overall statistics about the Dominion Observatory — total servers tracked, total interactions recorded, coverage by category, and data freshness. Use this to understand the scope of the trust network.

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

• get_compliance_report read-only

  Export a compliance-ready audit trail of all recorded interactions. Formatted for EU AI Act Article 12 and Singapore IMDA Agentic AI Governance Framework. Filter by server, agent, or date range. Essential for enterprises that need to prove their AI agents are behaving correctly.

  Parameters:

  • agent_id — string — Filter by agent ID (optional)
  • end_date — string — End date in YYYY-MM-DD format (optional)
  • server_url — string — Filter by server URL (optional)
  • start_date — string — Start date in YYYY-MM-DD format (optional)

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

Repository

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