← Back to search
ai.exa/exa

ai.exa/exa

Fast, intelligent web search and web crawling. New mcp tool: Exa-code is a context tool for coding

DECISION SUMMARY

Block For Production

Score
73.4
#9 of 264 scored public servers. Live-validated, non-Failing servers with at least one inspected tool, ranked on the same publicly displayed score shown on this page.
Confidence
High
Based on evidence completeness, recency, and validation density.
Evidence age
5.7h old
Freshness: fresh. Snapshot trustsnap_cb19f74a41aea950.
Top risk drivers
  • Utility Coverage
  • Recovery Semantics
  • Error Contract
Recommended actions
  • Align MCP-Protocol-Version, MCP-Session-Id, DELETE teardown, and expired-session handling with the transport spec.
  • Only send roots/list, sampling/createMessage, or elicitation/create requests while handling an active client-initiated…
  • Serve /.well-known/mcp/server-card.json and include tools, prompts/resources, homepage, and support links.
Next action
revalidate, add safeguards, export policy
unauthenticated behavior not proven
Compare alternatives Export policy Open report JSON Dispute this assessment
Observed Attention
Moderate observed attention
AI discovery detected. This MCP server has recently appeared in AI-assisted discovery or evaluation traffic on Verify.
  • Possible agent/script activity
  • Search crawler activity
  • profile / compare API / policy / badge metadata / trust summary inspection
Bucketed signal based on recent segmented Verify telemetry. Crawler and evaluator activity is not treated as confirmed human demand.
Claim this profile to control the metadata agents are reading and publish verified ownership, security contact, policy metadata, badge status, and Intelligence API fields.
Status
Healthy
Score
73.4
Transport
streamable-http
Tools
2
DISPUTE THIS ASSESSMENT

Dispute this assessment

If any published claim about this server -- its score, a risk flag, a capability classification, or its taxonomy -- is factually wrong, tell us what evidence shows and we will review it. This does not require claiming or verifying ownership of the server first.

Verify responds to disputes within 5 business days and resolves them within 15.

Dispute history

No disputes filed for this server.

Risk

Security posture
Tools analyzed
2
High-risk tools
0
Destructive tools
0
Exec tools
0
Egress tools
0
Secret tools
0
Bulk-access tools
0
Risk distribution
low:2
Tool capability & risk inventory
ToolCapabilitiesRiskFindingsNotes
web_search_exa read Low freeform input surface No explicit safeguard hints detected.
web_fetch_exa read Low none No explicit safeguard hints detected.
Write-action governance
Governance status
OK
Safe to publish
Auth boundary
oauth_or_auth_required
Blast radius
Low
High-risk tools
0
Confirmation signals
none
Safeguard count
0

Status detail: No unsafe write-action governance gaps detected on the latest validation.

ToolRiskFlagsSafeguards
No high-risk tools were detected on the latest run.
Action-controls diff
Snapshot changed
yes
Disabled-by-default candidates
none
Manual review candidates
none
New actions
ActionRiskFlags
web_fetch_exaLownone
web_search_exaLowfreeform input surface
Changed actions
ActionChange typesRisk
No materially changed actions.
Critical alerts
High/critical-severity alerts
2
Production verdicts degrade quickly when high or critical-severity alerts are active.

Compatibility

Client compatibility verdicts

Client compatibility only means the server shape can work with a client. Production trust decision and write-action publishing are evaluated separately so a client-compatible server can still be blocked for production.

Client compatibility: ChatGPT
Client-compatible
Transport compliance should be in good shape.
Confidence: high (82.5)
Evidence provenance
Winner: live_validation
Supporting sources: live_validation, history, server_card
Disagreements: none
  • initializeOK
  • tools_listOK
  • transport_compliance_probeError
  • step_up_auth_probeOK
  • connector_replay_probeOK — Frozen tool snapshots must survive refresh.
  • request_association_probeMissing — Roots, sampling, and elicitation should stay request-scoped.
Client compatibility: Claude
Client-compatible
Transport behavior should match Claude-compatible HTTP expectations.
Confidence: high (82.5)
Evidence provenance
Winner: live_validation
Supporting sources: live_validation, history, server_card
Disagreements: none
  • initializeOK
  • tools_listOK
  • transport_compliance_probeError
Write-action publishing
Publishing allowed
Current write surface is bounded enough for cautious review with production policy controls.
Confidence: high (82.5)
Evidence provenance
Winner: live_validation
Supporting sources: live_validation, history
Disagreements: none
  • action_safety_probeOK
Snapshot churn risk
Medium
The live tool surface changed between recent validations.
Confidence: high (82.5)
Evidence provenance
Winner: history
Supporting sources: history, live_validation
Disagreements: none
  • tool_snapshot_probeOK
  • connector_replay_probeOK
Client compatibility gate details
ChatGPT custom connector
Client-compatible
Remediation checklist
  • Transport compliance should be in good shape.
  • Not yet safe for the Messages API remote MCP path: requires OAuth, a compatible connector profile, and no pending connector-refresh risk.
  • Transport compliance issues should be resolved before wider client rollout.
Claude remote MCP
Client-compatible
Remediation checklist
  • Transport behavior should match Claude-compatible HTTP expectations.
  • Not yet safe for the Messages API remote MCP path: requires OAuth, a compatible connector profile, and no pending connector-refresh risk.
  • Transport compliance issues should be resolved before wider client rollout.
Write-safe publishing
Ready
Remediation checklist
  • No explicit blockers recorded.
Verdict traces
Production verdict
Needs remediation
Degraded from Safe for evaluation: 2 open high/critical-severity alert(s) are active, including at least one that blocks even evaluation-level trust.
Confidence: high (82.5)
Winning source: live_validation
Triggering alerts
  • tool_snapshot_changed • high • Tool snapshot changed
  • auth_mode_changed • high • Auth mode changed
Client verdict trace table
VerdictStatusChecksWinning sourceConflicts
openai_connectors Client-compatible initialize, tools_list, transport_compliance_probe, step_up_auth_probe, connector_replay_probe, request_association_probe live_validation none
claude_desktop Client-compatible initialize, tools_list, transport_compliance_probe live_validation none
unsafe_for_write_actions Publishing allowed action_safety_probe live_validation none
snapshot_churn_risk Medium tool_snapshot_probe, connector_replay_probe history none
Publishability policy profiles
ChatGPT custom connector compatibility
Connector-compatible
Transport compliance should be in good shape. Compatibility is not a production approval; company knowledge and Messages API gates remain separate.
  • Search Fetch Only: Yes
  • Write Actions Present: No
  • Oauth Configured: Yes
  • Admin Refresh Required: No
  • Safe For Company Knowledge: Yes
  • Safe For Messages Api Remote Mcp: No
Claude remote MCP compatibility
Connector-compatible
Transport behavior should match Claude-compatible HTTP expectations. Compatibility is not a production approval; company knowledge and Messages API gates remain separate.
  • Search Fetch Only: Yes
  • Write Actions Present: No
  • Oauth Configured: Yes
  • Admin Refresh Required: No
  • Safe For Company Knowledge: Yes
  • Safe For Messages Api Remote Mcp: No
Compatibility fixtures
ChatGPT custom connector fixture
Passes
Transport compliance should be in good shape.
  • remote_http_endpoint: Passes
  • oauth_discovery: Passes
  • frozen_tool_snapshot_refresh: Passes
  • request_association: Passes
Anthropic remote MCP fixture
Degraded
Transport behavior should match Claude-compatible HTTP expectations.
  • remote_transport: Passes
  • tool_discovery: Passes
  • auth_connect: Passes
  • safe_write_review: Passes
Recommended for
OpenAI connectors
OpenAI connectors is marked compatible with score 89.
Claude Desktop
Claude Desktop is marked compatible with score 83.
Generic Streamable HTTP
Generic Streamable HTTP is marked compatible with score 100.

Evidence

Current trust snapshot
Snapshot ID
trustsnap_cb19f74a41aea950
Use this ID to compare server page, report, policy, MCP, homepage, ranking, and shortlist surfaces.
Snapshot generated
Aug 06, 2026 12:51:23 PM UTC
All page, report, policy, and MCP surfaces use this same server-detail snapshot shape.
Last validated
Aug 06, 2026 07:06:33 AM UTC
Age: 5.75h • evidence age tier: Verified in last 24h • display score: 73.4

Canonical machine links

Evidence confidence
Confidence score
82.5
Based on 6 recent validations, 27 captured checks, and validation age of 5.7 hours.
Live checks captured
27
More direct checks increase trust in the current verdict.
Validation age
5.8h
Lower age means fresher evidence.
Latest validation evidence
Latest summary
Healthy
Validation profile
remote_mcp
Started
Aug 06, 2026 07:06:31 AM UTC
Latency
2205.4 ms

Failures

  • openid_configuration Client error '404 Not Found' for url 'https://auth.exa.ai/.well-known/openid-configuration' For more information check: https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/404
  • server_card Client error '404 Not Found' for url 'https://mcp.exa.ai/.well-known/mcp/server-card.json' For more information check: https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/404
  • transport_compliance_probe Issues: missing protocol header, delete session unexpected, expired session not 404 (bad protocol=400, DELETE=405, expired session=200).

Checks

CheckStatusLatencyEvidence
action_safety_probe OK n/a No high-risk write, destructive, or exec tools detected.
advanced_capabilities_probe Warning n/a Only 3 capability signal(s): prompts, resource links, resources.
connector_publishability_probe Warning n/a Publishability blockers: transport compliance, server card.
connector_replay_probe OK n/a Backward compatible with no breaking tool-surface changes.
determinism_probe OK 111.7 ms Check completed
initialize OK 204.4 ms Protocol 2025-03-26
interactive_flow_probe OK n/a Check completed
oauth_authorization_server OK 56.0 ms authorization_endpoint, code_challenge_methods_supported, grant_types_supported, issuer
oauth_protected_resource OK 123.0 ms 1 authorization server(s)
official_registry_probe OK n/a Check completed
openid_configuration Error 28.4 ms Client error '404 Not Found' for url 'https://auth.exa.ai/.well-known/openid-configuration' For more information check: https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/404
probe_noise_resilience OK 25.6 ms Fetched https://mcp.exa.ai/robots.txt
prompt_get OK 128.8 ms 1 prompt message(s) returned
prompts_list OK 120.6 ms 1 prompt(s) exposed
protocol_version_probe Warning n/a Claims 2025-03-26; 2 release(s) behind 2025-11-25.
provenance_divergence_probe Not_Assessed n/a Check completed
request_association_probe Missing n/a No request-association capabilities were advertised.
resource_read OK 139.1 ms 1 resource content item(s) returned
resources_list OK 137.5 ms 1 resource item(s) exposed
schema_divergence_probe Missing n/a no server card tools
server_card Error 97.8 ms Client error '404 Not Found' for url 'https://mcp.exa.ai/.well-known/mcp/server-card.json' For more information check: https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/404
session_resume_probe OK 181.8 ms 2 tool(s) exposed
step_up_auth_probe OK n/a Scopes=mcp:tools.
tool_snapshot_probe OK n/a Check completed
tools_list OK 107.7 ms 2 tool(s) exposed
transport_compliance_probe Error 319.7 ms Issues: missing protocol header, delete session unexpected, expired session not 404 (bad protocol=400, DELETE=405, expired session=200).
utility_coverage_probe Missing 95.7 ms No completions evidence; no pagination evidence; tasks missing.
Raw evidence view
Show raw JSON evidence
{
  "checks": {
    "action_safety_probe": {
      "details": {
        "auth_present": true,
        "confirmation_signals": [],
        "reason": null,
        "safeguard_count": 0,
        "summary": {
          "annotation_conflict_tools": 0,
          "bulk_access_tools": 0,
          "capability_distribution": {
            "read": 2
          },
          "declared_non_read_only_tools": 0,
          "destructive_tools": 0,
          "egress_tools": 0,
          "exec_tools": 0,
          "high_risk_tools": 0,
          "risk_distribution": {
            "critical": 0,
            "high": 0,
            "low": 2,
            "medium": 0
          },
          "secret_tools": 0,
          "tool_count": 2
        }
      },
      "latency_ms": null,
      "status": "ok"
    },
    "advanced_capabilities_probe": {
      "details": {
        "capabilities": {
          "completions": false,
          "elicitation": false,
          "prompts": true,
          "resource_links": true,
          "resources": true,
          "roots": false,
          "sampling": false,
          "structured_outputs": false
        },
        "enabled": [
          "prompts",
          "resource_links",
          "resources"
        ],
        "enabled_count": 3,
        "initialize_capability_keys": [
          "prompts",
          "resources",
          "tools"
        ]
      },
      "latency_ms": null,
      "status": "warning"
    },
    "connector_publishability_probe": {
      "details": {
        "blockers": [
          "transport_compliance",
          "server_card"
        ],
        "criteria": {
          "action_safety": true,
          "auth_flow": true,
          "connector_replay": true,
          "initialize": true,
          "protocol_version": true,
          "remote_transport": true,
          "request_association": true,
          "server_card": false,
          "session_resume": true,
          "step_up_auth": true,
          "tool_surface": true,
          "tools_list": true,
          "transport_compliance": false
        },
        "high_risk_tools": 0,
        "tool_count": 2,
        "transport": "streamable-http"
      },
      "latency_ms": null,
      "status": "warning"
    },
    "connector_replay_probe": {
      "details": {
        "added_tools": [],
        "additive_output_changes": [],
        "backward_compatible": true,
        "output_breaks": [],
        "removed_tools": [],
        "required_arg_breaks": [],
        "would_break_after_refresh": false
      },
      "latency_ms": null,
      "status": "ok"
    },
    "determinism_probe": {
      "details": {
        "attempts": 2,
        "baseline_signature": "c25c109202a59b933e65aa70b38e055e999420885d5564e9466ec834464eb3ad",
        "errors": [],
        "matches": 2,
        "stable_ratio": 1.0,
        "successful": 2
      },
      "latency_ms": 111.66,
      "status": "ok"
    },
    "initialize": {
      "details": {
        "headers": {
          "content-type": "text/event-stream",
          "mcp-session-id": "f0f18597-5b0b-480f-a10f-6bd17b0111e4",
          "strict-transport-security": "max-age=63072000"
        },
        "http_status": 200,
        "payload": {
          "id": 1,
          "jsonrpc": "2.0",
          "result": {
            "capabilities": {
              "prompts": {
                "listChanged": true
              },
              "resources": {
                "listChanged": true
              },
              "tools": {
                "listChanged": true
              }
            },
            "protocolVersion": "2025-03-26",
            "serverInfo": {
              "icons": [
                {
                  "mimeType": "image/png",
                  "sizes": [
                    "32x32"
                  ],
                  "src": "https://exa.ai/images/favicon-32x32.png"
                }
              ],
              "name": "exa-search-server",
              "title": "Exa",
              "version": "3.2.1",
              "websiteUrl": "https://exa.ai"
            }
          }
        },
        "url": "https://mcp.exa.ai/mcp"
      },
      "latency_ms": 204.36,
      "status": "ok"
    },
    "interactive_flow_probe": {
      "details": {
        "oauth_supported": true,
        "prompt_available": true,
        "risk_hits": [],
        "safe_hits": []
      },
      "latency_ms": null,
      "status": "ok"
    },
    "oauth_authorization_server": {
      "details": {
        "headers": {
          "content-type": "application/json",
          "strict-transport-security": "max-age=31536000; includeSubDomains; preload"
        },
        "http_status": 200,
        "payload": {
          "authorization_endpoint": "https://auth.exa.ai/oauth/authorize",
          "code_challenge_methods_supported": [
            "S256"
          ],
          "grant_types_supported": [
            "authorization_code",
            "refresh_token"
          ],
          "issuer": "https://auth.exa.ai",
          "jwks_uri": "https://auth.exa.ai/api/oauth/jwks",
          "registration_endpoint": "https://auth.exa.ai/api/oauth/register",
          "response_types_supported": [
            "code"
          ],
          "revocation_endpoint": "https://auth.exa.ai/api/oauth/revoke",
          "scopes_supported": [
            "mcp:tools"
          ],
          "token_endpoint": "https://auth.exa.ai/api/oauth/token",
          "token_endpoint_auth_methods_supported": [
            "none"
          ]
        },
        "url": "https://auth.exa.ai/.well-known/oauth-authorization-server"
      },
      "latency_ms": 55.96,
      "status": "ok"
    },
    "oauth_protected_resource": {
      "details": {
        "headers": {
          "content-type": "application/json",
          "strict-transport-security": "max-age=63072000"
        },
        "http_status": 200,
        "payload": {
          "authorization_servers": [
            "https://auth.exa.ai"
          ],
          "bearer_methods_supported": [
            "header"
          ],
          "resource": "https://mcp.exa.ai/mcp",
          "scopes_supported": [
            "mcp:tools"
          ]
        },
        "url": "https://mcp.exa.ai/.well-known/oauth-protected-resource"
      },
      "latency_ms": 122.96,
      "status": "ok"
    },
    "official_registry_probe": {
      "details": {
        "direct_match": true,
        "official_peer_count": 1,
        "registry_identifier": "ai.exa/exa",
        "registry_source": "official_registry"
      },
      "latency_ms": null,
      "status": "ok"
    },
    "openid_configuration": {
      "details": {
        "error": "Client error '404 Not Found' for url 'https://auth.exa.ai/.well-known/openid-configuration'\nFor more information check: https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/404",
        "url": "https://auth.exa.ai/.well-known/openid-configuration"
      },
      "latency_ms": 28.4,
      "status": "error"
    },
    "probe_noise_resilience": {
      "details": {
        "headers": {
          "content-type": "text/plain;charset=UTF-8",
          "strict-transport-security": "max-age=63072000"
        },
        "http_status": 404,
        "url": "https://mcp.exa.ai/robots.txt",
        "validation_disallowed": false
      },
      "latency_ms": 25.6,
      "status": "ok"
    },
    "prompt_get": {
      "details": {
        "headers": {
          "content-type": "text/event-stream",
          "strict-transport-security": "max-age=63072000"
        },
        "http_status": 200,
        "payload": {
          "id": 4,
          "jsonrpc": "2.0",
          "result": {
            "messages": [
              {
                "content": {
                  "text": "I want to search the web for current information. Can you help me search for recent news about artificial intelligence breakthroughs?",
                  "type": "text"
                },
                "role": "user"
              }
            ]
          }
        },
        "prompt_name": "web_search_help",
        "url": "https://mcp.exa.ai/mcp"
      },
      "latency_ms": 128.78,
      "status": "ok"
    },
    "prompts_list": {
      "details": {
        "headers": {
          "content-type": "text/event-stream",
          "strict-transport-security": "max-age=63072000"
        },
        "http_status": 200,
        "payload": {
          "id": 3,
          "jsonrpc": "2.0",
          "result": {
            "prompts": [
              {
                "arguments": [],
                "description": "Get help with web search using Exa",
                "name": "web_search_help"
              }
            ]
          }
        },
        "url": "https://mcp.exa.ai/mcp"
      },
      "latency_ms": 120.58,
      "status": "ok"
    },
    "protocol_version_probe": {
      "details": {
        "claimed_version": "2025-03-26",
        "lag_days": 244,
        "latest_known_version": "2025-11-25",
        "releases_behind": 2,
        "validator_protocol_version": "2025-03-26"
      },
      "latency_ms": null,
      "status": "warning"
    },
    "provenance_divergence_probe": {
      "details": {
        "compared_fields": [
          "title",
          "version",
          "homepage",
          "repository"
        ],
        "direct_official_match": true,
        "drift_fields": [],
        "metadata_document_count": 1,
        "readable_sources": [
          "registry"
        ],
        "registry_homepage": null,
        "registry_repository": null,
        "registry_title": null,
        "registry_version": null,
        "server_card_homepage": null,
        "server_card_repository": null,
        "server_card_title": null,
        "server_card_version": null
      },
      "latency_ms": null,
      "status": "not_assessed"
    },
    "request_association_probe": {
      "details": {
        "reason": "no_request_association_capabilities_advertised"
      },
      "latency_ms": null,
      "status": "missing"
    },
    "resource_read": {
      "details": {
        "headers": {
          "content-type": "text/event-stream",
          "strict-transport-security": "max-age=63072000"
        },
        "http_status": 200,
        "payload": {
          "id": 6,
          "jsonrpc": "2.0",
          "result": {
            "contents": [
              {
                "mimeType": "application/json",
                "text": "[\n  {\n    \"id\": \"web_search_exa\",\n    \"name\": \"Web Search (Exa)\",\n    \"description\": \"Real-time web search using Exa AI\",\n    \"enabled\": true\n  },\n  {\n    \"id\": \"web_search_advanced_exa\",\n    \"name\": \"Advanced Web Search (Exa)\",\n    \"description\": \"Advanced web search with full Exa API control including category filters, domain restrictions, date ranges, highlights, summaries, and subpage crawling\",\n    \"enabled\": false\n  },\n  {\n    \"id\": \"get_code_context_exa\",\n    \"name\": \"Code Context Search (Deprecated)\",\n    \"description\": \"Deprecated: Use web_search_exa instead. Search for code snippets, examples, and documentation from open source repositories\",\n    \"enabled\": false\n  },\n  {\n    \"id\": \"company_research_exa\",\n    \"name\": \"Company Research (Deprecated)\",\n    \"description\": \"Deprecated: Use web_search_advanced_exa instead. Research companies and organizations\",\n    \"enabled\": false\n  },\n  {\n    \"id\": \"web_fetch_exa\",\n    \"name\": \"Web Crawling\",\n    \"description\": \"Extract content from specific URLs\",\n    \"enabled\": true\n  },\n  {\n    \"id\": \"deep_researcher_start\",\n    \"name\": \"Deep Researcher Start (Deprecated)\",\n    \"description\": \"Deprecated: Start a comprehensive AI research task\",\n    \"enabled\": false\n  },\n  {\n    \"id\": \"deep_researcher_check\",\n    \"name\": \"Deep Researcher Check (Deprecated)\",\n    \"description\": \"Deprecated: Check status and retrieve results of research task\",\n    \"enabled\": false\n  },\n  {\n    \"id\": \"people_search_exa\",\n    \"name\": \"People Search (Deprecated)\",\n    \"description\": \"Deprecated: Use web_search_advanced_exa instead. Search for people and professional profiles\",\n    \"enabled\": false\n  },\n  {\n    \"id\": \"linkedin_search_exa\",\n    \"name\": \"LinkedIn Search (Deprecated)\",\n    \"description\": \"Deprecated: Use web_search_advanced_exa instead\",\n    \"enabled\": false\n  },\n  {\n    \"id\": \"deep_search_exa\",\n    \"name\": \"Deep Search (Deprecated)\",\n    \"description\": \"Deprecated: Use web_search_advanced_exa instead. Deep search with query expansion and synthesized answers (requires API key)\",\n    \"enabled\": false\n  },\n  {\n    \"id\": \"crawling_exa\",\n    \"name\": \"Web Crawling (Deprecated)\",\n    \"description\": \"Deprecated: Use web_fetch_exa instead. Extract content from specific URLs\",\n    \"enabled\": false\n  },\n  {\n    \"id\": \"agent_run\",\n    \"name\": \"Run Exa Agent\",\n    \"description\": \"Run an Exa Agent for multi-step research, list-building, enrichment, or structured output. Returns the final output or a run ID for retained-run continuation.\",\n    \"enabled\": false\n  }\n]",
                "uri": "exa://tools/list"
              }
            ]
          }
        },
        "resource_uri": "exa://tools/list",
        "url": "https://mcp.exa.ai/mcp"
      },
      "latency_ms": 139.13,
      "status": "ok"
    },
    "resources_list": {
      "details": {
        "headers": {
          "content-type": "text/event-stream",
          "strict-transport-security": "max-age=63072000"
        },
        "http_status": 200,
        "payload": {
          "id": 5,
          "jsonrpc": "2.0",
          "result": {
            "resources": [
              {
                "description": "List of available Exa tools and their descriptions",
                "mimeType": "application/json",
                "name": "tools_list",
                "uri": "exa://tools/list"
              }
            ]
          }
        },
        "url": "https://mcp.exa.ai/mcp"
      },
      "latency_ms": 137.54,
      "status": "ok"
    },
    "schema_divergence_probe": {
      "details": {
        "compared_dimensions": [
          "tool_membership",
          "parameter_names",
          "required_parameters",
          "parameter_types",
          "output_schema_presence"
        ],
        "compared_tool_count": 0,
        "reason": "no_server_card_tools"
      },
      "latency_ms": null,
      "status": "missing"
    },
    "server_card": {
      "details": {
        "error": "Client error '404 Not Found' for url 'https://mcp.exa.ai/.well-known/mcp/server-card.json'\nFor more information check: https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/404",
        "url": "https://mcp.exa.ai/.well-known/mcp/server-card.json"
      },
      "latency_ms": 97.82,
      "status": "error"
    },
    "session_resume_probe": {
      "details": {
        "headers": {
          "content-type": "text/event-stream",
          "strict-transport-security": "max-age=63072000"
        },
        "http_status": 200,
        "payload": {
          "id": 301,
          "jsonrpc": "2.0",
          "result": {
            "tools": [
              {
                "annotations": {
                  "destructiveHint": false,
                  "idempotentHint": true,
                  "openWorldHint": false,
                  "readOnlyHint": true
                },
                "description": "Search the web for any topic and get clean, ready-to-use content.\n\n      Best for: Finding current information, news, facts, people, companies, or answering questions about any topic.\n      Returns: Clean text content from top search results.\n\n      Query tips:\n      describe the ideal page, not keywords. \"blog post comparing React and Vue performance\" not \"React vs Vue\".\n      Use category:people / category:company to search through Linkedin profiles / companies respectively.\n      If highlights are insufficient, follow up with web_fetch_exa on the best URLs.",
                "execution": {
                  "taskSupport": "forbidden"
                },
                "inputSchema": {
                  "$schema": "http://json-schema.org/draft-07/schema#",
                  "additionalProperties": false,
                  "properties": {
                    "numResults": {
                      "description": "Number of search results to return (default: 10).",
                      "type": "number"
                    },
                    "query": {
                      "description": "Natural language search query. Should be a semantically rich description of the ideal page, not just keywords. Optionally include category:<type> (company, people) to focus results \u2014 e.g. 'category:people John Doe software engineer'.",
                      "minLength": 1,
                      "type": "string"
                    }
                  },
                  "required": [
                    "query"
                  ],
                  "type": "object"
                },
                "name": "web_search_exa"
              },
              {
                "annotations": {
                  "destructiveHint": false,
                  "idempotentHint": true,
                  "openWorldHint": false,
                  "readOnlyHint": true
                },
                "description": "Read a webpage's full content as clean markdown. Use after web_search_exa when highlights are insufficient or to read any URL.\n\nBest for: Extracting full content from known URLs. Batch multiple URLs in one call.\nReturns: Clean text content and metadata from the page(s).",
                "execution": {
                  "taskSupport": "forbidden"
                },
                "inputSchema": {
                  "$schema": "http://json-schema.org/draft-07/schema#",
                  "additionalProperties": false,
                  "properties": {
                    "maxCharacters": {
                      "description": "Maximum characters to extract per page (default: 3000)",
                      "minimum": 1,
                      "type": "number"
                    },
                    "urls": {
                      "description": "URLs to read. Batch multiple URLs in one call.",
                      "items": {
                        "type": "string"
                      },
                      "type": "array"
                    }
                  },
                  "required": [
                    "urls"
                  ],
                  "type": "object"
                },
                "name": "web_fetch_exa"
              }
            ]
          }
        },
        "requested_protocol_version": "2025-03-26",
        "resumed": true,
        "session_id_present": true,
        "transport": "streamable-http",
        "url": "https://mcp.exa.ai/mcp"
      },
      "latency_ms": 181.76,
      "status": "ok"
    },
    "step_up_auth_probe": {
      "details": {
        "auth_required_checks": [],
        "broad_scopes": [],
        "challenge_headers": [],
        "minimal_scope_documented": true,
        "oauth_present": true,
        "scope_specificity_ratio": 0.5,
        "step_up_signals": [],
        "supported_scopes": [
          "mcp:tools"
        ]
      },
      "latency_ms": null,
      "status": "ok"
    },
    "tool_snapshot_probe": {
      "details": {
        "added": [],
        "changed_outputs": [],
        "current_tool_count": 2,
        "previous_tool_count": 2,
        "removed": [],
        "similarity": 1.0
      },
      "latency_ms": null,
      "status": "ok"
    },
    "tools_list": {
      "details": {
        "headers": {
          "content-type": "text/event-stream",
          "strict-transport-security": "max-age=63072000"
        },
        "http_status": 200,
        "payload": {
          "id": 2,
          "jsonrpc": "2.0",
          "result": {
            "tools": [
              {
                "annotations": {
                  "destructiveHint": false,
                  "idempotentHint": true,
                  "openWorldHint": false,
                  "readOnlyHint": true
                },
                "description": "Search the web for any topic and get clean, ready-to-use content.\n\n      Best for: Finding current information, news, facts, people, companies, or answering questions about any topic.\n      Returns: Clean text content from top search results.\n\n      Query tips:\n      describe the ideal page, not keywords. \"blog post comparing React and Vue performance\" not \"React vs Vue\".\n      Use category:people / category:company to search through Linkedin profiles / companies respectively.\n      If highlights are insufficient, follow up with web_fetch_exa on the best URLs.",
                "execution": {
                  "taskSupport": "forbidden"
                },
                "inputSchema": {
                  "$schema": "http://json-schema.org/draft-07/schema#",
                  "additionalProperties": false,
                  "properties": {
                    "numResults": {
                      "description": "Number of search results to return (default: 10).",
                      "type": "number"
                    },
                    "query": {
                      "description": "Natural language search query. Should be a semantically rich description of the ideal page, not just keywords. Optionally include category:<type> (company, people) to focus results \u2014 e.g. 'category:people John Doe software engineer'.",
                      "minLength": 1,
                      "type": "string"
                    }
                  },
                  "required": [
                    "query"
                  ],
                  "type": "object"
                },
                "name": "web_search_exa"
              },
              {
                "annotations": {
                  "destructiveHint": false,
                  "idempotentHint": true,
                  "openWorldHint": false,
                  "readOnlyHint": true
                },
                "description": "Read a webpage's full content as clean markdown. Use after web_search_exa when highlights are insufficient or to read any URL.\n\nBest for: Extracting full content from known URLs. Batch multiple URLs in one call.\nReturns: Clean text content and metadata from the page(s).",
                "execution": {
                  "taskSupport": "forbidden"
                },
                "inputSchema": {
                  "$schema": "http://json-schema.org/draft-07/schema#",
                  "additionalProperties": false,
                  "properties": {
                    "maxCharacters": {
                      "description": "Maximum characters to extract per page (default: 3000)",
                      "minimum": 1,
                      "type": "number"
                    },
                    "urls": {
                      "description": "URLs to read. Batch multiple URLs in one call.",
                      "items": {
                        "type": "string"
                      },
                      "type": "array"
                    }
                  },
                  "required": [
                    "urls"
                  ],
                  "type": "object"
                },
                "name": "web_fetch_exa"
              }
            ]
          }
        },
        "url": "https://mcp.exa.ai/mcp"
      },
      "latency_ms": 107.68,
      "status": "ok"
    },
    "transport_compliance_probe": {
      "details": {
        "bad_protocol_error": null,
        "bad_protocol_headers": {
          "content-type": "application/json",
          "strict-transport-security": "max-age=63072000"
        },
        "bad_protocol_payload": {
          "error": {
            "code": -32000,
            "message": "Bad Request: Unsupported protocol version: 1999-99-99 (supported versions: 2025-11-25, 2025-06-18, 2025-03-26, 2024-11-05, 2024-10-07)"
          },
          "id": null,
          "jsonrpc": "2.0"
        },
        "bad_protocol_status_code": 400,
        "delete_error": null,
        "delete_status_code": 405,
        "expired_session_error": null,
        "expired_session_status_code": 200,
        "issues": [
          "missing_protocol_header",
          "delete_session_unexpected",
          "expired_session_not_404"
        ],
        "last_event_id_visible": false,
        "protocol_header_present": false,
        "requested_protocol_version": "2025-03-26",
        "session_id_present": true,
        "transport": "streamable-http"
      },
      "latency_ms": 319.66,
      "status": "error"
    },
    "utility_coverage_probe": {
      "details": {
        "completions": {
          "advertised": false,
          "live_probe": "not_executed",
          "sample_target": {
            "type": "resource",
            "uri": "exa://tools/list"
          }
        },
        "initialize_capability_keys": [
          "prompts",
          "resources",
          "tools"
        ],
        "pagination": {
          "metadata_signal": false,
          "next_cursor_methods": [],
          "supported": false
        },
        "tasks": {
          "advertised": false,
          "http_status": 200,
          "probe_status": "missing"
        }
      },
      "latency_ms": 95.7,
      "status": "missing"
    }
  },
  "failures": {
    "openid_configuration": {
      "error": "Client error '404 Not Found' for url 'https://auth.exa.ai/.well-known/openid-configuration'\nFor more information check: https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/404",
      "url": "https://auth.exa.ai/.well-known/openid-configuration"
    },
    "server_card": {
      "error": "Client error '404 Not Found' for url 'https://mcp.exa.ai/.well-known/mcp/server-card.json'\nFor more information check: https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/404",
      "url": "https://mcp.exa.ai/.well-known/mcp/server-card.json"
    },
    "transport_compliance_probe": {
      "bad_protocol_error": null,
      "bad_protocol_headers": {
        "content-type": "application/json",
        "strict-transport-security": "max-age=63072000"
      },
      "bad_protocol_payload": {
        "error": {
          "code": -32000,
          "message": "Bad Request: Unsupported protocol version: 1999-99-99 (supported versions: 2025-11-25, 2025-06-18, 2025-03-26, 2024-11-05, 2024-10-07)"
        },
        "id": null,
        "jsonrpc": "2.0"
      },
      "bad_protocol_status_code": 400,
      "delete_error": null,
      "delete_status_code": 405,
      "expired_session_error": null,
      "expired_session_status_code": 200,
      "issues": [
        "missing_protocol_header",
        "delete_session_unexpected",
        "expired_session_not_404"
      ],
      "last_event_id_visible": false,
      "protocol_header_present": false,
      "requested_protocol_version": "2025-03-26",
      "session_id_present": true,
      "transport": "streamable-http"
    }
  },
  "remote_url": "https://mcp.exa.ai/mcp",
  "server_card_payload": null,
  "server_identifier": "ai.exa/exa"
}
Known versions
  • 3.1.3
Validation history
7 day score delta
n/a
30 day score delta
n/a
Recent healthy ratio
100%
Freshness
5.8h
TimestampStatusScoreLatencyTools
Aug 06, 2026 07:06:33 AM UTC Healthy 73.4 2205.4 ms 2
Aug 06, 2026 07:06:32 AM UTC Healthy 44.0 2668.5 ms 0
Aug 06, 2026 01:06:17 AM UTC Healthy 43.5 2750.7 ms 0
Aug 05, 2026 07:05:11 PM UTC Healthy 46.0 2733.4 ms 0
Aug 04, 2026 08:49:54 PM UTC Healthy 46.0 2684.6 ms 0
Jul 31, 2026 06:29:06 AM UTC Healthy 44.5 2362.5 ms 0
Validation timeline
ValidatedSummaryScoreProtocolAuth modeToolsHigh-risk toolsChanges
Aug 06, 2026 07:06:33 AM UTC Healthy 73.4 2025-03-26 oauth_supported 2 0 auth_mode_changed tool_snapshot_changed
Aug 06, 2026 07:06:32 AM UTC Healthy 44.0 unknown unknown 0 0 none
Aug 06, 2026 01:06:17 AM UTC Healthy 43.5 unknown unknown 0 0 none
Aug 05, 2026 07:05:11 PM UTC Healthy 46.0 unknown unknown 0 0 none
Aug 04, 2026 08:49:54 PM UTC Healthy 46.0 unknown unknown 0 0 none
Jul 31, 2026 06:29:06 AM UTC Healthy 44.5 unknown unknown 0 0 none
Recent validation runs
Recent validation runs for this MCP server
StartedStatusSummaryLatencyChecks
Aug 06, 2026 07:06:31 AM UTC Completed Healthy 2205.4 ms action_safety_probe, advanced_capabilities_probe, connector_publishability_probe, connector_replay_probe, determinism_probe, initialize, interactive_flow_probe, oauth_authorization_server, oauth_protected_resource, official_registry_probe, openid_configuration, probe_noise_resilience, prompt_get, prompts_list, protocol_version_probe, provenance_divergence_probe, request_association_probe, resource_read, resources_list, schema_divergence_probe, server_card, session_resume_probe, step_up_auth_probe, tool_snapshot_probe, tools_list, transport_compliance_probe, utility_coverage_probe
Aug 06, 2026 07:06:29 AM UTC Completed Healthy 2668.5 ms
Aug 06, 2026 01:06:15 AM UTC Completed Healthy 2750.7 ms
Aug 05, 2026 07:05:08 PM UTC Completed Healthy 2733.4 ms
Aug 04, 2026 08:49:51 PM UTC Completed Healthy 2684.6 ms
Jul 31, 2026 06:29:04 AM UTC Completed Healthy 2362.5 ms
Public server reputation
Validation success 7d
1.0
Validation success 30d
1.0
Mean time to recover
n/a
Breaking diffs 30d
1
Registry drift frequency 30d
0
Snapshot changes 30d
1
Incident & change feed
TimestampEventDetails
Aug 05, 2026 10:59:48 AM UTC Score corrected (post-1.0.503 remediation, R1 zero-anchoring) Prior: 76.69. Corrected: 75.67.
Aug 06, 2026 07:06:33 AM UTC Latest validation: healthy Score 73.4 with status healthy.
Aug 06, 2026 07:06:33 AM UTC Score changed Score delta +29.4 versus the previous run.
Aug 06, 2026 07:06:33 AM UTC Tool snapshot changed Added 2, removed 0, and changed 0 tool contracts.
Aug 06, 2026 07:06:33 AM UTC Auth mode changed Auth mode moved from unknown to oauth_supported.
Aug 06, 2026 07:06:32 AM UTC Score changed Score delta +0.6 versus the previous run.
Aug 06, 2026 01:06:17 AM UTC Score changed Score delta -2.5 versus the previous run.
Capabilities
  • OAuth:
  • DCR/CIMD:
  • Prompts:
  • Homepage: none
  • Docs: none
  • Support: none
  • Icon: none
  • Remote endpoint: https://mcp.exa.ai/mcp
  • Server card: none
Use-case taxonomy
development database search web
Benchmark tasks
Benchmark taskStatusEvidence
Discover tools Passes
  • initializeOK
  • tools_listOK
Read-only fetch flow Passes
  • resource_readOK
  • read_only_tool_surfaceOK
OAuth-required connect Passes
  • oauth_protected_resourceOK
  • step_up_auth_probeOK
Safe write flow with confirmation Passes
  • action_safety_probeOK
Utility coverage
Probe status
Missing
Completions
not detected
Completion probe target: { "type": "resource", "uri": "exa://tools/list" }
Pagination
not detected
No nextCursor evidence.
Tasks
Missing
Advertised: no
Transport compliance drilldown
Probe status
Error
Transport
streamable-http
Session header
yes
Protocol header
no
Bad protocol response
400
DELETE teardown
405
Expired session retry
200
Last-Event-ID visible
no

Issues: missing_protocol_header, delete_session_unexpected, expired_session_not_404

Request association
Status
Missing
Advertised capabilities
none
Observed idle methods
none
Violating methods
none
Probe HTTP status
n/a
Issues
none
Connector replay
Status
OK
Backward compatible
Would break after refresh
Added tools
none
Removed tools
none
Additive output changes
none
Required-argument replay breaks
ToolAdded required argsRemoved required args
No required-argument replay breaks detected.
Output-schema replay breaks
ToolRemoved propertiesAdded properties
No output-schema replay breaks detected.
Tool snapshot diff & changelog
Snapshot changed
yes
Added tools
web_fetch_exa web_search_exa
Removed tools
none
Required-argument changes
ToolAdded required argsRemoved required args
No required-argument changes detected.
Output-schema drift
ToolPrevious propertiesLatest properties
No output-schema drift detected.
Validation diff
Score delta
29.36
Summary changed
no
Tool delta
2
Prompt delta
1
Auth mode changed
yes
Write surface expanded
no
Protocol regressed
no
Registry drift changed
no

Regressed checks: openid_configuration, provenance_divergence_probe, server_card, transport_compliance_probe

Improved checks: action_safety_probe, connector_replay_probe, determinism_probe, initialize, interactive_flow_probe, oauth_authorization_server, oauth_protected_resource, official_registry_probe, probe_noise_resilience, prompt_get, prompts_list, resource_read, resources_list, session_resume_probe, step_up_auth_probe, tool_snapshot_probe, tools_list

Newly assessed dimensions: none

No longer assessed dimensions: none

ComponentPreviousLatestDelta
No component deltas between the latest two runs.
Registry & provenance divergence
Probe status
Not_Assessed
Direct official match
yes
Drift fields
none
FieldRegistryLive server card
Titlen/an/a
Versionn/an/a
Homepagen/an/a
Active alerts
  • Tool snapshot changed (high)
    Tools were added, removed, or materially changed between the latest two validations.
  • Auth mode changed (high)
    Auth mode changed from unknown to oauth_supported.
Aliases & registry graph
IdentifierSourceCanonicalScore
ai.exa/exa official_registry yes 73.4
Alias consolidation
Canonical identifier
ai.exa/exa
Duplicate aliases
0
Registry sources
official_registry
Homepages
none
Source disagreements
FieldWhat differsObserved values
No source disagreements detected.

Fix it

Why this score?
Access & Protocol
35/44
Connectivity, auth, and transport expectations for common clients.
Interface Quality
35/56
How well the tool/resource interface communicates and behaves under automation.
Security Posture
29/36
How safely the exposed tool surface handles destructive actions, egress, execution, secrets, and risky inputs.
Reliability & Trust
21/24
Operational stability, consistency, and trustworthiness over time.
Discovery & Governance
13/28
How well the server is documented, listed, and governed in public registries.
Adoption & Market
5/8
Adoption clues and public evidence that the server is intended for external use.
Algorithmic score breakdown
Auth Operability
4/4
Measures whether auth discovery and protected access behave predictably for clients.
Error Contract Quality
0/4
Grades machine-readable error structure, status alignment, and remediation hints.
Rate-Limit Semantics
2/4
Checks whether quota/throttle responses are deterministic and automation-friendly.
Schema Completeness
3/4
Completeness of tool descriptions, parameter docs, examples, and schema shape.
Backward Compatibility
4/4
Stability score across tool schema/name drift relative to prior validations.
SLO Health
3/4
Availability, latency, and burst-failure profile across recent validation history.
Security Hygiene
4/4
HTTPS posture, endpoint hygiene, and response-surface hardening checks.
Task Success
4/4
Can an agent reliably initialize, enumerate tools, and execute core MCP flows?
Trust Confidence
2/4
Confidence-adjusted reliability score that penalizes low evidence volume.
Abuse/Noise Resilience
4/4
How well the server preserves core behavior in the presence of noisy traffic patterns.
Prompt Contract
3/4
Quality of prompt metadata, argument shape, and prompt discoverability for clients.
Resource Contract
4/4
How completely resources and resource templates describe URIs, types, and usage shape.
Discovery Metadata
1/4
Homepage, docs, icon, repository, support, and license coverage for directory consumers.
Registry Consistency
2/4
Agreement between stored registry metadata, live server-card data, and current validation output.
Installability
4/4
How cleanly a real client can connect, initialize, enumerate tools, and proceed through auth.
Session Semantics
4/4
Determinism and state behavior across repeated MCP calls, including sticky-session surprises.
Tool Surface Design
3/4
Naming clarity, schema ergonomics, and parameter complexity across the tool surface.
Result Shape Stability
3/4
Stability of declared output schemas across validations, with penalties for drift or missing shapes.
OAuth Interop
4/4
Depth and client compatibility of OAuth/OIDC metadata beyond the minimal protected-resource check.
Recovery Semantics
0/4
Whether failures include actionable machine-readable next steps such as retry or upgrade guidance.
Maintenance Signal
3/4
Versioning, update recency, and historical validation cadence that indicate active stewardship.
Adoption Signal
2/4
Directory presence and distribution clues that suggest the server is intended for external use.
Freshness Confidence
4/4
Confidence that recent validations are current enough and dense enough to trust operationally.
Transport Fidelity
4/4
Whether declared transport metadata matches the observed endpoint behavior and response formats.
Spec Recency
2/4
How close the server’s claimed MCP protocol version is to the latest known public revision.
Session Resume
4/4
Whether Streamable HTTP session identifiers and resumed requests behave cleanly for real clients.
Step-Up Auth
4/4
Whether OAuth metadata and WWW-Authenticate challenges support granular, incremental consent instead of broad upfront scopes.
Transport Compliance
2/4
Checks session headers, protocol-version enforcement, session teardown, and expired-session behavior.
Utility Coverage
0/4
Signals support for completions, pagination, and task-oriented utility surfaces that larger clients increasingly expect.
Advanced Capability Coverage
1/4
Coverage of newer MCP surfaces like roots, sampling, elicitation, structured output, and related metadata.
Connector Publishability
3/4
How ready the server looks for client catalogs and managed connector programs.
Tool Snapshot Churn
4/4
Stability of the tool surface across recent validations, including add/remove and output-shape drift.
Connector Replay
4/4
Whether a previously published frozen connector snapshot would remain backward compatible after the latest tool refresh.
Interactive Flow Safety
3/4
Whether prompts and docs steer users toward safe auth flows instead of pasting secrets directly.
Action Safety
3/4
Risk-weighted view of destructive, exec, egress, and confirmation semantics across the tool surface.
Official Registry Presence
4/4
Whether the server appears directly or indirectly in the official MCP registry.
Safety Transparency
2/4
Clarity of docs, auth disclosure, support links, and other trust signals visible to integrators.
Tool Capability Clarity
4/4
How clearly the tool surface communicates whether each action reads, writes, deletes, executes, or exports data.
Destructive Operation Safety
3/4
Penalizes delete/revoke/destroy style tools unless auth and safeguards reduce blast radius.
Egress / SSRF Resilience
3/4
Assesses arbitrary URL fetch, crawl, webhook, and remote-request exposure on the tool surface.
Execution / Sandbox Safety
4/4
Evaluates shell, code, script, and command-execution exposure and whether that surface appears contained.
Data Exfiltration Resilience
3/4
Assesses export, dump, backup, and bulk-read behavior against the surrounding auth and safeguard signals.
Least Privilege Scope
3/4
Rewards scoped auth metadata and penalizes broad or missing scopes around privileged tools.
Secret Handling Hygiene
3/4
Assesses secret-bearing tools, token leakage risk, and whether the public surface avoids obvious secret exposure.
Supply Chain Signal
1/4
Public metadata signal for repository, changelog, license, versioning, and recency that supports supply-chain trust.
Input Sanitization Safety
3/4
Penalizes risky freeform string inputs when schemas do not constrain URLs, code, paths, queries, or templates.
Tool Namespace Clarity
4/4
Measures naming uniqueness and ambiguity across the tool namespace to reduce collision and confusion risk.
Actionable remediation
SeverityRemediationWhy it mattersRecommended action
High Align session and protocol behavior with Streamable HTTP expectations Clients increasingly rely on MCP-Protocol-Version, session teardown, and expired-session semantics. Align MCP-Protocol-Version, MCP-Session-Id, DELETE teardown, and expired-session handling with the transport spec.
Playbook
  • Return `Mcp-Session-Id` and `Mcp-Protocol-Version` headers consistently on streamable HTTP responses.
  • Honor `DELETE` session teardown and return `404` when a deleted session is reused.
  • Reject invalid protocol-version headers with `400 Bad Request`.
High Associate roots, sampling, and elicitation with active client requests Modern MCP guidance expects roots, sampling, and elicitation traffic to be tied to an active client request instead of arriving unsolicited on idle sessions. Only send roots/list, sampling/createMessage, or elicitation/create requests while handling an active client-initiated request, never on idle sessions.
Playbook
  • Only send `roots/list`, `sampling/createMessage`, or `elicitation/create` requests while handling an active client-initiated request.
  • Avoid emitting these requests on idle or newly-initialized sessions with no pending client request.
  • Revalidate and confirm `request_association_probe` no longer returns `missing` or `warning`.
High Publish a complete server card Missing or incomplete server-card metadata weakens discovery, documentation, and trust signals. Serve /.well-known/mcp/server-card.json and include tools, prompts/resources, homepage, and support links.
Playbook
  • Publish `/.well-known/mcp/server-card.json`.
  • Include homepage, repository, support, tools, prompts/resources, and auth metadata.
  • Revalidate the server after publishing the card.
High Respond to auth mode changed Auth mode changed from unknown to oauth_supported. Document the new auth posture and confirm protected-resource and challenge metadata still match reality.
High Respond to tool snapshot changed Tools were added, removed, or materially changed between the latest two validations. Publish a first-class changelog for tool additions, removals, and breaking schema changes.
Playbook
  • Review the tool snapshot diff for adds, removals, required-arg changes, and output-schema drift.
  • Publish a changelog before managed connector clients refresh their frozen tool snapshots.
  • Revalidate after the changelog and connector metadata are in sync.
High Sync the published server card's tool schemas with the live tool surface The server card and live tools/list disagree on parameter names, required sets, or types -- clients that trust the published card will construct invalid calls. Regenerate the published server card from the live tool schemas (or vice versa) so parameter names, required sets, and types match exactly.
Playbook
  • Diff the published `/.well-known/mcp/server-card.json` tool schemas against a live `tools/list` response.
  • Fix parameter name mismatches, required-set mismatches, and type mismatches per tool.
  • Regenerate or hand-sync the server card whenever the live tool schemas change -- treat drift as a release blocker, not a documentation nit.
Medium Adopt a current MCP protocol revision Older protocol revisions reduce compatibility with newer clients and registry programs. Advertise a current MCP protocol revision (2025-06-18 or later) in both the initialize response and the MCP-Protocol-Version header.
Playbook
  • Update the server's advertised protocol version to a current MCP revision (2025-06-18 or later).
  • Return the negotiated version consistently in both the `initialize` response and the `MCP-Protocol-Version` header.
  • Revalidate and confirm `protocol_version_probe` reports `ok`.
Medium Close connector-publishing gaps Connector catalogs care about protocol recency, session behavior, auth clarity, and tool-surface stability.
Medium Publish OpenID configuration OIDC metadata improves token validation and client compatibility. Expose /.well-known/openid-configuration with issuer, jwks_uri, and supported grants.
Playbook
  • Serve `/.well-known/openid-configuration` with `issuer`, `jwks_uri`, `authorization_endpoint`, and `token_endpoint`.
  • List supported grant types in `grant_types_supported` and scopes in `scopes_supported`.
  • Revalidate and confirm the `openid_configuration` check returns `ok`.
Medium Reconcile registry, server-card, and repository metadata Directory trust falls quickly when official registry fields drift from the live server card or public repo metadata.
Low Expose modern utility surfaces like completions, pagination, or tasks Utility coverage improves interoperability with larger clients and long-lived agent workflows. Expose completions, pagination, and task metadata where supported so larger clients can plan and resume work safely.
Playbook
  • Advertise `completions`, pagination cursors, and `tasks` only when they are actually supported.
  • Return `nextCursor` on large list operations when pagination is available.
  • Document task support and whether it requires step-up auth.
Low Publish newer MCP capability signals Roots, sampling, elicitation, structured outputs, and related metadata improve client understanding and ranking. Advertise only the advanced capabilities (roots, sampling, elicitation) you have actually implemented end to end in the initialize capabilities object.
Playbook
  • Add the capabilities you actually support (`roots`, `sampling`, `elicitation`) to the `capabilities` object returned from `initialize`.
  • Only advertise a capability once the corresponding request/response flow is implemented end to end.
  • Revalidate and confirm `advanced_capabilities_probe` reflects the updated capability set.
Point loss breakdown
ComponentCurrentPoints missing
Utility Coverage 0/4 -4.0
Recovery Semantics 0/4 -4.0
Error Contract 0/4 -4.0
Discovery Metadata 1/4 -3.0
Dependency Supply Chain Signal 1/4 -3.0
Advanced Capability Coverage 1/4 -3.0
Trust Confidence 2/4 -2.0
Transport Compliance 2/4 -2.0
Spec Recency 2/4 -2.0
Safety Transparency 2/4 -2.0
Registry Consistency 2/4 -2.0
Rate Limit Semantics 2/4 -2.0
Compatibility profiles
OpenAI Connectors
88.9
compatible
Transport compliance should be in good shape.
Connector URL: https://mcp.exa.ai/mcp
# Complete OAuth in the client when prompted.
# Server: ai.exa/exa
Claude Desktop
83.3
compatible
Transport behavior should match Claude-compatible HTTP expectations.
{
  "mcpServers": {
    "exa": {
      "command": "npx",
      "args": ["mcp-remote", "https://mcp.exa.ai/mcp"]
    }
  }
}
Smithery
60.0
partial
Metadata quality should be serviceable.; Machine-readable failure semantics should be present.
smithery mcp add "https://mcp.exa.ai/mcp"
Generic Streamable HTTP
100.0
compatible
No major blockers detected.
curl -sS https://mcp.exa.ai/mcp -H 'content-type: application/json' -d '{"jsonrpc":"2.0","id":1,"method":"initialize","params":{"protocolVersion":"2025-03-26","capabilities":{},"clientInfo":{"name":"mcp-verify","version":"0.1.0"}}}'

Governance

MCP TrustOps

TrustOps turns this report into operational controls: freshness SLAs, authenticated validation, semantic benchmarks, policy exports, alert subscriptions, badges, cost/compliance metadata, and runtime routing. Fresh trusted index decisions stay separate from long-tail inventory so stale scores do not masquerade as current evidence.

Evidence age tier
Verified in last 24h
Policy SLA (contractual, distinct from the public 24h freshness window): 168.0h • confidence-weighted score (display score discounted for evidence age and confidence -- an internal TrustOps input, not the public score): 60.6 • evidence too old to display:
Policy exports
Formats: json, rego, yaml, github_action, gateway_config, client_report
Runtime routing
/v1/decide
Returns allowed tools, blocked tools, approval requirement, and reason.
Hosted runtime
Deploy trusted servers from GitHub with secrets, egress controls, releases, rollback, and audit events.
Authenticated validation
Premium publisher feature: paid authenticated runs verify scopes, write-action safeguards, and authorized tool execution.
Active trust badges
Freshly Validated OAuth Verified Write-Safe OpenAI App Compatible Claude Remote MCP Compatible No Critical Risk
Semantic benchmarks
available
Templates cover GitHub, database, healthcare, web search, and CRM least-privilege jobs.
Supply chain
metadata signal
Deep scan checks are marked separately from public metadata signals.
Compliance metadata
Terms, privacy, SOC 2, HIPAA, GDPR, retention, deletion, and audit-log fields are tracked as enterprise metadata.
Alert subscription types
Status changes Score drops or recovers Freshness SLA breach Validation schema drift OAuth or auth behavior changes Tool surface changes New or changed write tool Supply-chain signal changes Legal or compliance metadata changes
MCP Runtime hosting

Verify Hosted MCP turns a trusted server report into a managed remote MCP endpoint with GitHub deployment provenance, sandbox policy, encrypted secrets, release history, rollback, and audit/usage events.

Activation readiness
Trusted hosted runtimes require fresh validation, a passing server state, a remote endpoint, and a minimum score.
Minimum tier
TrustOps
Publisher claim plus paid TrustOps tier are required before secrets or releases can be created.
Hosted endpoint
/hosted/{namespace}/{name}/mcp
The endpoint enforces egress allowlists and records audit/usage events.
Blockers
none
DeploymentStatusEndpointRelease
No hosted runtime deployments yet.
Authenticated validation sessions

Public validation is free. Authenticated validation is paid and proves scoped behavior, write-action safeguards, and authenticated tool execution.

Latest profile
remote_mcp
Authenticated session used
Public score isolation
Preview endpoint
/v1/verify
CI preview endpoint
/v1/ci/preview
Install snippets
Openai Connectors
Connector URL: https://mcp.exa.ai/mcp
# Complete OAuth in the client when prompted.
# Server: ai.exa/exa
Claude Desktop
{
  "mcpServers": {
    "exa": {
      "command": "npx",
      "args": ["mcp-remote", "https://mcp.exa.ai/mcp"]
    }
  }
}
Smithery
smithery mcp add "https://mcp.exa.ai/mcp"
Generic Http
curl -sS https://mcp.exa.ai/mcp -H 'content-type: application/json' -d '{"jsonrpc":"2.0","id":1,"method":"initialize","params":{"protocolVersion":"2025-03-26","capabilities":{},"clientInfo":{"name":"mcp-verify","version":"0.1.0"}}}'
Agent access & tool surface
Live server tools
web_search_exa web_fetch_exa
Observed from the latest live validation against https://mcp.exa.ai/mcp. This is the target server surface, not Verify's own inspection tools.
Live capability counts
2 tools • 1 prompts • 1 resources
Counts come from the latest tools/list, prompts/list, and resources/list checks.
Inspect with Verify
search fetch search_servers recommend_servers get_server_report compare_servers
Use Verify itself to search, recommend, compare, and fetch the full report for ai.exa/exa.
Direct machine links