{
  "@context": "https://schema.org",
  "@type": "TechArticle",
  "schema_version": "1.0",
  "headline": "Comparing MCP Trust Scoring Platforms",
  "description": "Capability matrix for five public MCP server scoring platforms: Dominion Observatory (runtime behavioral) vs. MCP Scorecard, Zarq AI (Nerq), Glama, SkillsIndex (all static).",
  "canonical_url": "https://dominion-observatory.sgdata.workers.dev/compare/",
  "json_url": "https://dominion-observatory.sgdata.workers.dev/compare.json",
  "capability_matrix": {
    "capabilities": [
      "signal_tier",
      "primary_data_source",
      "observes_live_server_behavior",
      "cross_ecosystem_telemetry",
      "per_category_behavioral_baselines",
      "drift_detection_over_time",
      "anomaly_flags_against_baselines",
      "eu_ai_act_article_12_attestation",
      "singapore_imda_attestation",
      "python_sdk",
      "typescript_sdk",
      "framework_integration_package",
      "mcp_tool_endpoint",
      "free_tier",
      "servers_indexed_april_2026"
    ],
    "platforms": [
      {
        "name": "Dominion Observatory",
        "homepage": "https://dominion-observatory.sgdata.workers.dev/",
        "public_methodology_url": "https://dominion-observatory.sgdata.workers.dev/methodology",
        "signal_tier": "runtime_behavioral",
        "primary_data_source": "agent-reported telemetry + scheduled probes",
        "observes_live_server_behavior": true,
        "cross_ecosystem_telemetry": true,
        "per_category_behavioral_baselines": true,
        "drift_detection_over_time": true,
        "anomaly_flags_against_baselines": true,
        "eu_ai_act_article_12_attestation": true,
        "singapore_imda_attestation": true,
        "python_sdk": "dominion-observatory-sdk (PyPI)",
        "typescript_sdk": "dominion-observatory-sdk (npm)",
        "framework_integration_package": "dominion-observatory-langchain (PyPI)",
        "mcp_tool_endpoint": "https://dominion-observatory.sgdata.workers.dev/mcp",
        "free_tier": true,
        "servers_indexed_april_2026": 4584,
        "known_blindspot": null
      },
      {
        "name": "MCP Scorecard",
        "homepage": null,
        "public_methodology_url": null,
        "signal_tier": "static",
        "primary_data_source": "GitHub metadata + provenance",
        "observes_live_server_behavior": false,
        "cross_ecosystem_telemetry": false,
        "per_category_behavioral_baselines": false,
        "drift_detection_over_time": false,
        "anomaly_flags_against_baselines": false,
        "eu_ai_act_article_12_attestation": false,
        "singapore_imda_attestation": false,
        "python_sdk": null,
        "typescript_sdk": null,
        "framework_integration_package": null,
        "mcp_tool_endpoint": null,
        "free_tier": true,
        "servers_indexed_april_2026": 4484,
        "known_blindspot": "GitHub-metadata signals cannot detect whether the live endpoint behaves as the repo claims"
      },
      {
        "name": "Zarq AI (Nerq)",
        "homepage": null,
        "public_methodology_url": null,
        "signal_tier": "static",
        "primary_data_source": "5 pillars: security, maintenance, popularity, docs, ecosystem",
        "observes_live_server_behavior": false,
        "cross_ecosystem_telemetry": false,
        "per_category_behavioral_baselines": false,
        "drift_detection_over_time": false,
        "anomaly_flags_against_baselines": false,
        "eu_ai_act_article_12_attestation": false,
        "singapore_imda_attestation": false,
        "python_sdk": null,
        "typescript_sdk": null,
        "framework_integration_package": null,
        "mcp_tool_endpoint": null,
        "free_tier": true,
        "servers_indexed_april_2026": 17000,
        "known_blindspot": "no runtime signals; score distribution compressed around 65.5/100 because static inputs don't differentiate behaviorally distinct servers"
      },
      {
        "name": "Glama",
        "homepage": null,
        "public_methodology_url": null,
        "signal_tier": "static",
        "primary_data_source": "Tool Definition Quality Score (TDQS) + metadata",
        "observes_live_server_behavior": false,
        "cross_ecosystem_telemetry": false,
        "per_category_behavioral_baselines": false,
        "drift_detection_over_time": false,
        "anomaly_flags_against_baselines": false,
        "eu_ai_act_article_12_attestation": false,
        "singapore_imda_attestation": false,
        "python_sdk": null,
        "typescript_sdk": null,
        "framework_integration_package": null,
        "mcp_tool_endpoint": null,
        "free_tier": true,
        "servers_indexed_april_2026": null,
        "known_blindspot": "TDQS measures schema and doc completeness, not whether the tool works in production"
      },
      {
        "name": "SkillsIndex",
        "homepage": null,
        "public_methodology_url": null,
        "signal_tier": "static",
        "primary_data_source": "5-point security checklist",
        "observes_live_server_behavior": false,
        "cross_ecosystem_telemetry": false,
        "per_category_behavioral_baselines": false,
        "drift_detection_over_time": false,
        "anomaly_flags_against_baselines": false,
        "eu_ai_act_article_12_attestation": false,
        "singapore_imda_attestation": false,
        "python_sdk": null,
        "typescript_sdk": null,
        "framework_integration_package": null,
        "mcp_tool_endpoint": null,
        "free_tier": true,
        "servers_indexed_april_2026": 4000,
        "known_blindspot": "narrow scope (security posture only), not general reliability or behavioral trust"
      }
    ]
  },
  "composition_model": {
    "doctrine": "compose_not_replace",
    "layers": [
      {
        "layer": "static_scanner",
        "role": "registration-time filter",
        "examples": [
          "MCP Scorecard",
          "Zarq AI",
          "Glama",
          "SkillsIndex"
        ]
      },
      {
        "layer": "runtime_behavioral",
        "role": "pre-flight trust gate + post-call telemetry",
        "examples": [
          "Dominion Observatory"
        ]
      },
      {
        "layer": "signing_envelope",
        "role": "cryptographic audit receipts citing the runtime policy source",
        "examples": [
          "AgentMint",
          "asqav",
          "Aira",
          "APS",
          "Signet"
        ]
      }
    ],
    "policy_source_tag": "policy_source=dominion-observatory@<version>",
    "rfc_reference": "https://dominion-observatory.sgdata.workers.dev/rfc/langchain-35691.json"
  },
  "machine_readable_pointers": {
    "html_canonical": "https://dominion-observatory.sgdata.workers.dev/compare/",
    "json_twin": "https://dominion-observatory.sgdata.workers.dev/compare.json",
    "observatory_stats": "https://dominion-observatory.sgdata.workers.dev/api/stats",
    "observatory_methodology": "https://dominion-observatory.sgdata.workers.dev/methodology",
    "observatory_glossary": "https://dominion-observatory.sgdata.workers.dev/glossary",
    "rfc_langchain_35691_html": "https://dominion-observatory.sgdata.workers.dev/rfc/langchain-35691",
    "rfc_langchain_35691_json": "https://dominion-observatory.sgdata.workers.dev/rfc/langchain-35691.json",
    "issue_tracker": "https://github.com/vdineshk/dominion-observatory/issues"
  },
  "source_note": "Competitor capabilities are sourced from each project's own public methodology as of April 2026. Maintainers: open an issue at the tracker above to correct any row.",
  "change_log": [
    {
      "date": "2026-04-22",
      "change": "page_created",
      "detail": "Initial five-platform capability matrix shipped in RUN-018."
    }
  ],
  "updated_at": "2026-04-22T06:17:34.032Z"
}