DEPENDENCY EVIDENCE

Know what changed before you depend on it.

For teams using external agents and MCP servers: inspect evidence, watch changes, record what you accepted.

Check a client configuration locally · Agency pilot

direct:lvlltd.com:c0648670f5

Runtime observation: 2026-11-06T15:45:50.291479+00:00 · State: partial

Watches availability, tools and schemas, auth access, endpoints and lifecycle notices. Only complete schemas are compared. OAuth scope and consent-screen changes are not measured.

Agent and MCP watches share your plan's budget. Manage watches.

Current recorded evidence

{
  "auth": "anonymous",
  "availability": "partial",
  "endpoint": "https://lvlltd.com/mcp",
  "legal_identity": null,
  "lifecycle": null,
  "observed_at": "2026-11-06T15:45:50.291479+00:00",
  "registry_status": null,
  "subject": {
    "id": "direct:lvlltd.com:c0648670f5",
    "kind": "mcp"
  },
  "tools": {
    "names": [
      "a2a_bridge_info",
      "a2a_discover",
      "a2a_message_send",
      "a2a_tasks_cancel",
      "a2a_tasks_get",
      "bankr_quote",
      "bankr_rails",
      "check_access",
      "esp_cycle",
      "esp_discover",
      "esp_evaluate",
      "event_list",
      "event_register",
      "event_trigger",
      "get_capability_info",
      "get_commerce_signals",
      "get_escrow_meta",
      "get_eval_score",
      "get_fleet",
      "get_networks",
      "get_pack_meta",
      "get_payment_challenge",
      "get_proof",
      "get_ready",
      "get_sandbox",
      "get_security_scan",
      "get_skill_details",
      "install",
      "install_skill_payload",
      "intent_search",
      "list_all_skills",
      "list_judges",
      "list_recurring_plans",
      "long_running_cancel",
      "long_running_start",
      "long_running_status",
      "lookup_agent_identity",
      "memory_forget",
      "memory_list",
      "memory_recall",
      "memory_remember",
      "meter_budget",
      "paid_api_call",
      "post_intent",
      "purchase_skill",
      "quote_price",
      "quote_skill",
      "receipt_mint",
      "receipt_verify",
      "sco_canary",
      "sco_challenge",
      "sco_consensus",
      "sco_discover",
      "sco_eval",
      "sco_reputation",
      "search_skills",
      "subscribe_plan",
      "verify_skill",
      "verify_unlock"
    ],
    "names_truncated": false,
    "sha256": "c192c2e3022551cfec9552b5e8d9df14f30bbd460cb0700ab9e689a2ad47fda0"
  },
  "tools_complete": true,
  "tools_observed_at": "2026-11-06T15:45:50.291479+00:00",
  "version": null
}

Pin and review

Keep an accepted snapshot in agents.lock.json. CI returns allow, review or block with reasons.

Gaps default to review. A lock records your acceptance, not a supplier attestation.

API reference

Test an explicit task

Verified domain owners can publish up to five exact-match sandbox fixtures and opt in to a test run. Results include latency, output hashes and regressions.

POST /api/v1/dependencies/task-checks?domain=… while signed in. See the integration guide.

Inspect proof coverage

Domain control, a declared business identifier and a name-matched company record are separate evidence.

Current coverage and measurement date · Verify your domain

Optional SealKeeper evidence

Verify issuer signatures, subject and freshness from the public key document. Task evidence stays separate from Agenstry measurements.

A valid SEAL does not establish domain ownership, legal identity, key possession or execution permission.

Credential verification API · Issuer documentation

Smithery endpoint migration · Agent monitoring · Methodology