{"generated_at":"2026-10-14T20:07:22+00:00","item":{"added_at":"2026-08-05","analysis_method":"independent_manual_review_2026-10-10","assessment_status":"provisional","auth_posture":"auth_required","category":"Web Connector","confidence":"medium","description":"GatherHQ connects Claude to Emma for customer research. Plan studies, retrieve findings, explore personas, and create research-based content. Real-study plans and interview guides require approval before fielding.","did":"3d5477ac-f07b-46db-b08d-2e520b18ada5","documentation_url":"https://share.gatherhq.com/mcp","editorial_description":{"extended_description":"The connector exposes a conversational ask-Emma tool that carries context across follow-ups and approvals. A GatherHQ account connection is required.","schema_version":1,"summary":"GatherHQ connects Claude to Emma for customer research.\nPlan studies, retrieve findings, explore personas, and create research-based content.\nReal-study plans and interview guides require approval before fielding."},"homepage_url":"","icon":"","id":"3d5477ac-f07b-46db-b08d-2e520b18ada5","installs":0,"last_scanned":"","license":"","long_description":"The connector exposes a conversational ask-Emma tool that carries context across follow-ups and approvals. A GatherHQ account connection is required.","manual_assessment":"high","manual_review":{"all_findings":[{"evidence":"The description says ask_emma can scope and launch real studies, recruit participants, pull results and create reports, decks, posts and ad creatives, while only one generic tool is inventoried.","reasoning":"Concrete paid, outbound and mutating actions cannot be reviewed or policy-gated individually from the stored surface.","risk_type":"generic_gateway","severity":"high","title":"One generic agent tool hides a full research workflow","triggering_tools_or_skills":["ask_emma"]}],"assessment_status":"provisional","auth_and_scopes":"Authentication connects a GatherHQ research workspace, but the single generic tool's action schema, approval enforcement and study-budget permissions are not captured.","blocked_missing_evidence":null,"confidence":"medium","displayed_risk":"high","evidence_completeness":"partial","manual_overall_risk":"high","official_sources":[{"accessed_at":"2026-10-10T20:24:37Z","url_or_path":"plutosecurity/plutonium@b749c8dc43536c794409be9805edd48063f045e2:app/data/conn/claudesec/3d5477ac-f07b-46db-b08d-2e520b18ada5.json","what_it_proves":"Candidate PR #125 record identity (name \"GatherHQ\", publisher, type web_connector) and 1 stored tool names; no tool descriptions stored"},{"accessed_at":"2026-09-22T10:46:07Z","url_or_path":"https://api.anthropic.com/api/directory/servers?verified_tier=anthropic,partner,community (snapshot claude-directory-all-tiers-2026-09-22.json)","what_it_proves":"Official Claude Directory listing https://claude.ai/directory/3d5477ac-f07b-46db-b08d-2e520b18ada5: complete tool_names inventory (1 tools), remote streamable-http endpoint https://share.gatherhq.com/api/mcp, auth_posture auth_required, directory tier community"}],"primary_reason":"GatherHQ collapses real-study launch, participant recruitment and content creation into an opaque agent gateway.","provisional_risk":null,"publication_recommendation":"fix","requires_runtime_verification":true,"review_key":"claude:3d5477ac-f07b-46db-b08d-2e520b18ada5","reviewed_at":"2026-10-10","reviewer_notes":"Snapshot the ask_emma input schema and prove that study launch/recruitment require a distinct human confirmation; determine whether launches spend credits or send participant invitations.","risk_domains":{"arbitrary_network_access":"false","authz_ambiguity":"true","browser_control":"false","credential_or_admin_actions":"false","deletes_data":"unknown","deployment_or_infrastructure":"false","dynamic_hidden_actions":"true","filesystem_or_repo_access":"false","forwards_data_to_third_party":"true","modifies_data":"true","persistent_credentials":"true","prompt_injection_exposure":"true","provenance_uncertainty":"true","reads_private_data":"true","runs_code_or_queries":"false","sends_messages_as_user":"unknown","spends_or_moves_money":"unknown"},"tools_or_skills_reviewed":1,"tools_or_skills_total":1},"name":"GatherHQ","needs_runtime_verification":true,"plutonium_url":"https://plutonium.pluto.security/detail.html?planet=claudesec&did=3d5477ac-f07b-46db-b08d-2e520b18ada5&utm_source=plutonium_analysis_skill&utm_medium=claude_skill&utm_campaign=connector_risk_assessment","privacy_policy_url":"https://www.gatherhq.com/privacy-policy","provisional_risk":null,"publication_recommendation":"fix","publisher":"GatherHQ","publisher_url":"https://gatherhq.com/","remote_url":"https://share.gatherhq.com/api/mcp","repository_url":"","risk":"high","risk_severity":"high","security_risks":[{"description":"This connector can execute code, commands, queries, or dynamically selected actions. Review findings: Concrete paid, outbound and mutating actions cannot be reviewed or policy-gated individually from the stored surface.","evidence":"One generic agent tool hides a full research workflow: The description says ask_emma can scope and launch real studies, recruit participants, pull results and create reports, decks, posts and ad creatives, while only one generic tool is inventoried.","manual_risk_type":"generic_gateway","remediation":{"block_tool_categories":["code_execution"],"steps":["Run only in a sandbox with least-privilege credentials and strict resource limits.","Disable generic execution gateways unless their reachable action set is fully understood.","Require approval before execution against production systems or sensitive data."]},"risk_type":"runs_code","severity":"high","title":"One generic agent tool hides a full research workflow","triggering_tools":["ask_emma"]}],"signature_status":"unknown","source_code_reviewed":false,"support_url":"success@gatherhq.com","tags":["network_access","oauth_required","runs_code","web_connector"],"tools":[{"description":"","name":"ask_emma","risk":{"category":"code_execution","level":"high","recommendation":"Follow the connector-level remediation and require approval for consequential actions.","why":"This connector can execute code, commands, queries, or dynamically selected actions. Review findings: Concrete paid, outbound and mutating actions cannot be reviewed or policy-gated individually from the stored surface."}}],"tools_count":1,"transport":"streamable-http","type":"web_connector","url":"https://claude.ai/directory/3d5477ac-f07b-46db-b08d-2e520b18ada5","uuid":"53eacf82-53c7-5884-88e3-42260e0f08c4","verification_tier":"community","version":""},"query_key":"gatherhq gatherhq","schema_version":1,"status":"match"}