generated: '2026-08-29' method: derived source: >- openapi/relevance-ai-openapi.json (521 paths / 566 operations) bound against the MCP tool names the provider publishes in RelevanceAI/cc-plugin and RelevanceAI/agent-skills provider: Relevance AI providerId: relevance-ai description: >- Binds each Relevance AI MCP tool to the REST operation(s) that back it, so an agent holding only the MCP tool list can recover the real parameter and request-body shapes from the OpenAPI. Mapping is by name and documented semantics: the live MCP tools/list is OAuth-gated (401 invalid_token anonymously), so no inputSchema was read and confidence is set honestly. surfaces: openapi: file: openapi/relevance-ai-openapi.json source_url: https://api-f1db6c.stack.tryrelevance.com/latest/openapi_schema.json gated: false operations: 566 mcp: url: https://mcp.relevanceai.com/ gated: true gate: OAuth 2.0 (authorization_code + PKCE); anonymous tools/list returns 401 invalid_token tools_named_by_provider_docs: 125 graphql: present: false naming_note: >- The REST API calls a Tool a "studio" (ListStudios, TriggerStudio, CloneStudio, BulkDeleteStudios) while the MCP layer and the product docs call it a "tool". Both names are the provider's own; the crosswalk is the only place the two vocabularies are reconciled. crosswalk: - tool: relevance_list_agents category: agents rest: [ListAgents] binding: direct confidence: high - tool: relevance_get_agent category: agents rest: [GetAgent] binding: direct confidence: high - tool: relevance_upsert_agent category: agents rest: [UpsertAgent] binding: direct confidence: high - tool: relevance_update_agent category: agents rest: [UpsertAgent] binding: direct confidence: medium note: The REST surface exposes upsert only; the provider's own skill warns that a save wipes omitted fields. - tool: relevance_save_agent_draft category: agents rest: [SaveAgentAsDraft] binding: direct confidence: high - tool: relevance_publish_agent category: agents rest: [PublishAgent] binding: direct confidence: high - tool: relevance_restore_agent_version category: agents rest: [RestoreAgentVersion] binding: direct confidence: high - tool: relevance_list_agent_versions category: agents rest: [ListAgentVersions] binding: direct confidence: high - tool: relevance_get_agent_version category: agents rest: [GetAgentVersion] binding: direct confidence: high - tool: relevance_get_agent_tools category: agents rest: [GetAgentsTools] binding: direct confidence: high - tool: relevance_attach_tools_to_agent category: agents rest: [GetAgent, SaveAgentAsDraft, PublishAgent] binding: composite confidence: high note: >- The provider's skill documents this tool as performing fetch, merge, save, publish and action-ID retrieval in one call — a composite over three REST operations. - tool: relevance_detach_tools_from_agent category: agents rest: [GetAgent, SaveAgentAsDraft, PublishAgent] binding: composite confidence: medium - tool: relevance_read_agent_system_prompt category: agents rest: [GetAgent] binding: projection confidence: medium - tool: relevance_edit_agent_system_prompt category: agents rest: [GetAgent, SaveAgentAsDraft] binding: composite confidence: medium - tool: relevance_search_agent_system_prompt category: agents rest: [GetAgent] binding: projection confidence: low - tool: relevance_trigger_agent category: agents rest: [TriggerAgent] binding: direct confidence: high - tool: relevance_trigger_agent_sync category: agents rest: [TriggerAgent] binding: direct confidence: medium note: Synchronous variant, documented 120-second timeout. - tool: relevance_poll_agent_result category: agents rest: [ListAgentTaskView, GetAgentConversationMetadata] binding: composite confidence: low note: Async polling helper; which REST read it wraps is not stated in the docs. - tool: relevance_list_agent_tasks category: agents rest: [ListTaskGroups, ListTasksForGroup] binding: composite confidence: medium - tool: relevance_list_agent_task_messages category: agents rest: [ListAgentTaskView] binding: direct confidence: medium - tool: relevance_get_agent_task_metadata category: agents rest: [GetAgentConversationMetadata] binding: direct confidence: medium - tool: relevance_get_agent_task_summary category: agents rest: [GetAgentConversationMetadata] binding: projection confidence: low - tool: relevance_list_agent_triggers category: triggers rest: [ListAgentTriggerGroupsSummaries, ListAgentActiveTriggerGroupsOverview] binding: composite confidence: medium - tool: relevance_create_trigger category: triggers rest: [UpsertAgentScheduledTriggersGroup, CreateAgentScheduledTriggersItem] binding: composite confidence: medium - tool: relevance_delete_trigger category: triggers rest: [CancelAgentScheduledTriggerItem] binding: direct confidence: low note: REST exposes cancel rather than delete for scheduled trigger items. - tool: relevance_meeting_bot category: agents rest: [GetMeetingAgentConnectionDetails] binding: direct confidence: low - tool: relevance_list_tools category: tools rest: [ListStudios, ListStudiosView] binding: direct confidence: high - tool: relevance_search_tools category: tools rest: [ListStudiosView, ListStudiosFoldersAwareView] binding: direct confidence: medium - tool: relevance_get_tool category: tools rest: [GetStudio] binding: direct confidence: high - tool: relevance_upsert_tool category: tools rest: [UpsertToolVersion] binding: direct confidence: medium - tool: relevance_create_tool category: tools rest: [UpsertToolVersion] binding: direct confidence: medium - tool: relevance_update_tool category: tools rest: [UpsertToolVersion] binding: direct confidence: medium - tool: relevance_publish_tool category: tools rest: [PublishTool] binding: direct confidence: high - tool: relevance_restore_tool_version category: tools rest: [RestoreToolVersion] binding: direct confidence: high - tool: relevance_list_tool_versions category: tools rest: [ListToolVersions] binding: direct confidence: high - tool: relevance_get_tool_version category: tools rest: [GetToolVersion] binding: direct confidence: high - tool: relevance_bulk_delete_tools category: tools rest: [BulkDeleteStudios] binding: direct confidence: high - tool: relevance_trigger_tool category: tools rest: [TriggerStudio] binding: direct confidence: high - tool: relevance_run_tool category: tools rest: [TriggerStudio] binding: direct confidence: high - tool: relevance_trigger_tool_async category: tools rest: [TriggerStudioAsync] binding: direct confidence: high - tool: relevance_poll_tool_result category: tools rest: [StudioAsyncPoll] binding: direct confidence: high - tool: relevance_get_latest_tool_run category: tools rest: [ListStudiosRunHistory] binding: direct confidence: medium - tool: relevance_search_public_tools category: tools rest: [ListPublicStudios] binding: direct confidence: medium - tool: relevance_clone_public_tool category: tools rest: [CloneStudio] binding: direct confidence: medium - tool: relevance_list_transformations category: tools rest: [ListStudioTransformations] binding: direct confidence: high - tool: relevance_search_transformations category: tools rest: [ListStudioTransformations] binding: direct confidence: high - tool: relevance_create_tool_from_transformation category: tools rest: [GetStudioTransformation, UpsertToolVersion] binding: composite confidence: medium - tool: relevance_add_tool_step category: tools rest: [GetStudio, UpsertToolVersion] binding: composite confidence: medium note: Tool steps are edited inside the studio document, so every step mutation is a read-modify-upsert. - tool: relevance_update_tool_step category: tools rest: [GetStudio, UpsertToolVersion] binding: composite confidence: medium - tool: relevance_move_tool_step category: tools rest: [GetStudio, UpsertToolVersion] binding: composite confidence: medium - tool: relevance_remove_tool_step category: tools rest: [GetStudio, UpsertToolVersion] binding: composite confidence: medium - tool: relevance_set_tool_output category: tools rest: [GenerateToolDisplayOutput, ApplyToolDisplayOutput] binding: composite confidence: low - tool: relevance_list_workforces category: workforces rest: [ListWorkforces] binding: direct confidence: high - tool: relevance_get_workforce category: workforces rest: [GetWorkforce] binding: direct confidence: high - tool: relevance_create_workforce category: workforces rest: [SaveWorkforce, UpsertWorkforceVersion] binding: composite confidence: medium - tool: relevance_update_workforce category: workforces rest: [SaveWorkforce] binding: direct confidence: medium - tool: relevance_delete_workforce category: workforces rest: [DeleteWorkforce] binding: direct confidence: high - tool: relevance_publish_workforce category: workforces rest: [UpsertWorkforceVersion] binding: direct confidence: low - tool: relevance_trigger_workforce category: workforces rest: [TriggerWorkforce] binding: direct confidence: high - tool: relevance_list_workforce_tasks category: workforces rest: [ListWorkforceTasks] binding: direct confidence: high - tool: relevance_get_workforce_task_messages category: workforces rest: [ListWorkforceTaskMessages] binding: direct confidence: high - tool: relevance_get_workforce_task_metadata category: workforces rest: [GetWorkforceConversationMetadata] binding: direct confidence: high - tool: relevance_poll_workforce_task_result category: workforces rest: [GetWorkforceConversationMetadata, ListWorkforceTaskMessages] binding: composite confidence: medium - tool: relevance_list_workforce_versions category: workforces rest: [ListWorkforceVersions] binding: direct confidence: high - tool: relevance_list_knowledge_sets category: knowledge rest: [ListKnowledgeSets] binding: direct confidence: high - tool: relevance_get_knowledge_set category: knowledge rest: [GetKnowledgeSetMetadata] binding: direct confidence: high - tool: relevance_create_knowledge_set category: knowledge rest: [AddKnowledge] binding: direct confidence: medium note: The provider's skill states tables are created implicitly when the first row is added. - tool: relevance_update_knowledge_set category: knowledge rest: [UpdateKnowledgeSetMetadata] binding: direct confidence: high - tool: relevance_delete_knowledge_set category: knowledge rest: [DeleteKnowledgeSet] binding: direct confidence: high - tool: relevance_add_knowledge_rows category: knowledge rest: [AddKnowledge] binding: direct confidence: high - tool: relevance_list_knowledge_rows category: knowledge rest: [ListKnowledge] binding: direct confidence: high - tool: relevance_get_knowledge_row category: knowledge rest: [ListKnowledge] binding: projection confidence: medium - tool: relevance_update_knowledge_rows category: knowledge rest: [BulkUpdateKnowledge] binding: direct confidence: high - tool: relevance_delete_knowledge_rows category: knowledge rest: [DeleteKnowledge] binding: direct confidence: high - tool: relevance_export_knowledge category: knowledge rest: [KnowledgeExportTrigger] binding: direct confidence: high - tool: relevance_run_evaluation category: evals rest: [EvaluateResource] binding: direct confidence: high - tool: relevance_list_eval_batches category: evals rest: [ListEvalBatchByResourceNew] binding: direct confidence: high - tool: relevance_get_eval_batch_summary category: evals rest: [GetEvalBatchSummaryForResource] binding: direct confidence: high - tool: relevance_poll_eval_batch_result category: evals rest: [GetEvalBatchSummaryForResource] binding: direct confidence: medium - tool: relevance_cancel_eval_batch category: evals rest: [CancelEvalBatchForResource] binding: direct confidence: high - tool: relevance_delete_eval_batch category: evals rest: [DeleteEvalBatchForResource] binding: direct confidence: high - tool: relevance_list_eval_runs category: evals rest: [ListEvalRunsForResource] binding: direct confidence: high - tool: relevance_list_eval_checks category: evals rest: [ListEvalRulesForResource] binding: direct confidence: medium note: MCP calls them checks; REST calls them eval rules. - tool: relevance_create_eval_check category: evals rest: [CreateEvalRuleForResource] binding: direct confidence: medium - tool: relevance_update_eval_check category: evals rest: [UpdateEvalRuleForResource] binding: direct confidence: medium - tool: relevance_delete_eval_check category: evals rest: [DeleteEvalRuleForResource] binding: direct confidence: medium - tool: relevance_create_performance_dashboard category: analytics rest: [CreateObservabilityConfigV2] binding: direct confidence: low - tool: relevance_list_performance_dashboards category: analytics rest: [ListObservabilityConfigs] binding: direct confidence: low - tool: relevance_update_performance_dashboard category: analytics rest: [UpdateObservabilityConfigV2] binding: direct confidence: low - tool: relevance_delete_performance_dashboard category: analytics rest: [DeleteObservabilityConfigV2] binding: direct confidence: low - tool: relevance_get_performance_dashboard_timeseries category: analytics rest: [GetObservabilityEvalScoreTimeseriesForResource] binding: direct confidence: low - tool: relevance_get_analytics_usage_breakdown category: analytics rest: [GetAgentUsageBreakdown, GetToolUsageBreakdown, GetWorkforceUsageBreakdown] binding: composite confidence: medium - tool: relevance_get_analytics_usage_timeseries category: analytics rest: [GetAgentUsageTimeseries, GetToolUsageTimeseries, GetWorkforceUsageTimeseries] binding: composite confidence: medium - tool: relevance_get_analytics_agent_activity_timeseries category: analytics rest: [GetAgentUsageBreakdownTimeseries] binding: direct confidence: low - tool: relevance_get_analytics_errors category: analytics rest: [GetAgentErrors] binding: direct confidence: high - tool: relevance_get_analytics_resource_breakdown category: analytics rest: [GetAgentBreakdown, GetWorkforceBreakdown] binding: composite confidence: medium - tool: relevance_get_analytics_task_status_counts category: analytics rest: [GetTaskOpsTabCounts] binding: direct confidence: medium - tool: relevance_search_marketplace_listings category: marketplace rest: [ListPublishedMarketplaceListings] binding: direct confidence: high - tool: relevance_list_folders category: folders rest: [ListFolders] binding: direct confidence: high - tool: relevance_upsert_folder category: folders rest: [UpsertFolder] binding: direct confidence: high - tool: relevance_delete_folder category: folders rest: [DeleteFolder] binding: direct confidence: high - tool: relevance_get_project_info category: projects rest: [GetProjectsOrgInfo] binding: direct confidence: medium - tool: relevance_list_oauth_accounts category: auth rest: [ListOAuthTokenData] binding: direct confidence: medium - tool: relevance_submit_feedback category: support rest: [SubmitMessageFeedback] binding: direct confidence: medium - tool: relevance_raw_api category: raw rest: [] binding: passthrough confidence: high note: >- Escape hatch. The provider's skill documents it as issuing arbitrary REST calls (e.g. POST /knowledge/add, /knowledge/list, /knowledge/bulk_update, /knowledge/delete), so it can reach any of the 566 operations in the OpenAPI. No single operationId binds. - tool: relevance_api_request category: raw rest: [] binding: passthrough confidence: medium mcp_only: - tool: relevance_create_eval_test_case reason: No public REST operation for eval test cases in the harvested spec. - tool: relevance_get_eval_test_case reason: No public REST operation for eval test cases in the harvested spec. - tool: relevance_update_eval_test_case reason: No public REST operation for eval test cases in the harvested spec. - tool: relevance_delete_eval_test_case reason: No public REST operation for eval test cases in the harvested spec. - tool: relevance_list_eval_test_cases reason: No public REST operation for eval test cases in the harvested spec. - tool: relevance_add_eval_test_case_rule reason: No public REST operation for eval test-case rules in the harvested spec. - tool: relevance_remove_eval_test_case_rule reason: No public REST operation for eval test-case rules in the harvested spec. - tool: relevance_set_eval_test_case_simulation_config reason: No public REST operation for eval test-case simulation config in the harvested spec. - tool: relevance_create_eval_test_set reason: No public REST operation for eval test sets in the harvested spec. - tool: relevance_get_eval_test_set reason: No public REST operation for eval test sets in the harvested spec. - tool: relevance_update_eval_test_set reason: No public REST operation for eval test sets in the harvested spec. - tool: relevance_delete_eval_test_set reason: No public REST operation for eval test sets in the harvested spec. - tool: relevance_list_eval_test_sets reason: No public REST operation for eval test sets in the harvested spec. - tool: relevance_add_performance_dashboard_rule reason: Composite over observability config editing; no single REST operation. - tool: relevance_remove_performance_dashboard_rule reason: Composite over observability config editing; no single REST operation. - tool: relevance_list_performance_dashboard_runs reason: No matching REST operation identified in the harvested spec. rest_only_families: - family: organizations operations: 55 note: Org creation, membership, roles, subscriptions, usage alarms, data-retention config. - family: auth operations: 52 note: Users, keys, invites, OAuth account linking, SSO. Only account listing is surfaced via MCP. - family: billing operations: 22 note: Credit aggregation, subscriptions, downgrades, cancellation. Not exposed as MCP tools. - family: chat operations: 171 note: >- The largest REST family. Conversations, task views, streaming tokens, embed keys, bulk triggers and batch scheduling are mostly not reachable from the MCP tool list. - family: replicate / fal_ai / apify / inventor / alarms / user-memory operations: 20 note: Model-vendor passthrough and internal surfaces with no MCP counterpart. coverage: mcp_tools_named: 125 mcp_tools_mapped: 106 mcp_only: 16 passthrough: 2 rest_operations_total: 566 rest_operations_bound: 92 rest_operations_unbound: 474 note: >- rest_operations_bound counts distinct operationIds named at least once above. The large unbound remainder is genuine: the MCP server exposes a curated build/run surface, not the whole platform API.