generated: '2026-08-13' method: derived source: >- mcp/profound-mcp.yml (tool catalog from docs.tryprofound.com/mcp/capabilities/*) bound against openapi/_original/profound-openapi.json note: >- The live MCP tools/list is OAuth-gated (HTTP 401), so tools are mapped to REST operations by name and documented semantics rather than by comparing declared inputSchemas. Every operationId below was grepped verbatim out of the published OpenAPI. Confidence is set honestly per row; where a tool has both a v1 and v2 backing operation the v2 form is listed first because the docs describe the tool behaviour in v2 terms. surfaces: openapi: file: openapi/_original/profound-openapi.json title: External API version: 6551bd49db665ffb7f80780b3e150b8e8b780cc3 servers: [https://api.tryprofound.com] operations: 125 gated: false note: Spec is anonymously readable; calling it requires an Enterprise API key. mcp: url: https://mcp.tryprofound.com/mcp gated: true note: tools/list returns 401 invalid_token; OAuth 2.1 or a Bearer API key required. graphql: endpoint: null note: No GraphQL surface published; /graphql on the API host returns 404. crosswalk: - tool: whoami category: discovery rest: [] binding: mcp-only confidence: high note: >- No REST equivalent. Session identity/entitlement introspection is an MCP server concern; the REST API scopes by API key implicitly. - tool: list_organizations category: discovery rest: [list_organizations_v1_org_get] binding: rest confidence: high - tool: list_regions category: discovery rest: [get_regions_v1_org_regions_get, get_category_regions_v1_org_categories__category_id__regions_get] binding: rest confidence: high note: Fans out to org-level and category-scoped region listings. - tool: list_models category: discovery rest: [get_models_v1_org_models_get] binding: rest confidence: high - tool: list_categories category: discovery rest: [get_categories_v1_org_categories_get] binding: rest confidence: high - tool: list_domains category: discovery rest: [get_domains_v1_org_domains_get] binding: rest confidence: high - tool: list_tags category: discovery rest: [get_category_tags_v1_org_categories__category_id__tags_get] binding: rest confidence: high - tool: list_topics category: discovery rest: [get_category_topics_v1_org_categories__category_id__topics_get] binding: rest confidence: high - tool: list_prompts category: discovery rest: [get_category_prompts_v1_org_categories__category_id__prompts_get] binding: rest confidence: high note: Docs describe optional status, tag and topic filters, matching the query parameters on this operation. - tool: get_visibility_report category: brand-visibility rest: [query_visibility_v2_v2_reports_visibility_post, query_visibility_v1_reports_visibility_post] binding: rest confidence: high - tool: get_sentiment_report category: brand-visibility rest: [query_sentiment_v2_v2_reports_sentiment_post, query_sentiment_v1_reports_sentiment_post] binding: rest confidence: high - tool: get_citations_report category: brand-visibility rest: [query_citations_v2_v2_reports_citations_post, query_citations_v1_reports_citations_post] binding: rest confidence: high - tool: get_prompt_answers category: brand-visibility rest: [query_answers_v2_v2_prompts_answers_post, get_answers_v1_prompts_answers_post] binding: rest confidence: high - tool: get_factcheck_report category: brand-visibility rest: [query_scores_v2_reports_factcheck_post] binding: rest confidence: high - tool: get_factcheck_claims category: brand-visibility rest: [query_claims_v2_reports_factcheck_claims_post] binding: rest confidence: high - tool: get_referrals_report category: traffic rest: [get_referrals_report_v2_v2_reports_referrals_post, get_referrals_report_v1_v1_reports_referrals_post] binding: rest confidence: high - tool: get_bots_report category: traffic rest: [get_bots_report_v2_v2_reports_bots_post, get_bots_report_v1_v1_reports_bots_post] binding: rest confidence: high - tool: list_agents category: agents rest: [list_agents_v1_agents_get] binding: rest confidence: high - tool: get_agent category: agents rest: [get_agent_v1_agents__agent_id__get] binding: rest confidence: high - tool: run_agent category: agents rest: [run_agent_v1_agents__agent_id__runs_post] binding: rest confidence: high - tool: get_agent_run category: agents rest: [get_agent_run_v1_agents__agent_id__runs__run_id__get] binding: rest confidence: high - tool: list_agent_node_types category: agents rest: [list_node_types_v1_agents_node_types_get] binding: rest confidence: high - tool: get_agent_node_schema category: agents rest: [get_node_type_schema_v1_agents_node_types__node_type__schema_get] binding: rest confidence: high - tool: create_agent_definition category: agents rest: [create_agent_v1_agents_post] binding: rest confidence: medium note: >- The MCP tool adds a preview flag and a build-session id that the REST operation does not expose; the underlying create is the same operation. - tool: update_agent_definition category: agents rest: [update_agent_v1_agents__agent_id__patch] binding: rest confidence: medium note: Same preview/build-session divergence as create_agent_definition. - tool: publish_agent_definition category: agents rest: [publish_agent_v1_agents__agent_id__publish_post] binding: rest confidence: medium note: MCP wraps the publish with a preview plan step not present in REST. - tool: get_agent_definition category: agents rest: [get_agent_graph_v1_agents__agent_id__graph_get, get_agent_v1_agents__agent_id__get] binding: rest confidence: medium note: Reading a draft definition maps onto the agent and its graph. mcp_only: - tool: whoami reason: >- Session-scoped identity and entitlement introspection. No public REST operation returns the caller principal. - tool: start_agent_build_session reason: >- Server-side build-session orchestration that traces a multi-call agent build attempt. No REST operation creates or returns a build session. - tool: validate_agent_definition reason: >- Structural pre-check exposed only through MCP. The REST surface validates on publish; there is no standalone validate operation. - tool: list_agent_definition_templates reason: >- Pre-built Agent templates are not exposed by any published REST operation. rest_only: - capability: Knowledge bases operations: - list_knowledge_bases_v1_knowledge_bases_get - search_knowledge_base_v1_knowledge_bases__knowledge_base_id__search_post - add_document_v1_knowledge_bases__knowledge_base_id__documents_post - update_document_v1_knowledge_bases__knowledge_base_id__documents_put - delete_document_v1_knowledge_bases__knowledge_base_id__documents_delete - add_folder_v1_knowledge_bases__knowledge_base_id__folders_post - delete_folder_v1_knowledge_bases__knowledge_base_id__folders_delete note: >- The agent card and the published Agent Skill both advertise knowledge-base management, but the MCP capability pages list no knowledge-base tools. - capability: Projects and tasks operations: - list_projects_v1_projects_get - create_project_v1_projects_post - get_project_v1_projects__project_id__get - delete_project_v1_projects__project_id__delete - get_project_status_v1_projects__project_id__status_get - archive_project_v1_projects__project_id__archive_post - unarchive_project_v1_projects__project_id__unarchive_post - list_project_generations_v1_projects_generations_get - get_project_generation_status_v1_projects_generations__run_id__get - list_project_tasks_v1_projects__project_id__tasks_get - create_project_task_v1_projects__project_id__tasks_post - get_project_task_v1_projects__project_id__tasks__task_id__get - update_project_task_v1_projects__project_id__tasks__task_id__patch - delete_project_task_v1_projects__project_id__tasks__task_id__delete - update_project_task_status_v1_projects__project_id__tasks__task_id__status_post - capability: Documents operations: - list_documents_v1_documents_get - create_document_v1_documents_post - read_document_v1_documents__document_id__get - patch_document_v1_documents__document_id__patch - delete_document_v1_documents__document_id__delete - replace_document_content_v1_documents__document_id__content_post - capability: Shopping visibility operations: - query_shopping_brands_v2_v2_reports_shopping_brands_post - query_shopping_products_v2_v2_reports_shopping_products_post - query_shopping_merchants_v2_v2_reports_shopping_merchants_post - query_shopping_trigger_rate_v2_v2_reports_shopping_trigger_rate_post - shopping_visibility_v1_reports_shopping_visibility_post - shopping_item_visibility_v1_reports_shopping_item_visibility_post - shopping_merchant_distribution_v1_reports_shopping_merchant_distribution_post - shopping_merchant_visibility_by_brand_v1_reports_shopping_merchant_visibility_by_brand_post - shopping_merchant_by_items_v1_reports_shopping_merchant_by_items_post - shopping_all_items_with_merchants_v1_reports_shopping_all_items_with_merchants_post - shopping_trigger_rate_v1_reports_shopping_trigger_rate_post - shopping_triggered_prompts_v1_reports_shopping_triggered_prompts_post - shopping_triggered_topics_v1_reports_shopping_triggered_topics_post - shopping_merchant_share_v1_reports_shopping_merchant_share_post - shopping_product_merchant_urls_v1_reports_shopping_product_merchant_urls_post - shopping_executions_v1_reports_shopping_executions_post note: The entire Shopping product line is REST-only — no MCP tool exposes it. - capability: Accuracy / FactCheck detail operations: - accuracy_overview_v1_reports_accuracy_overview_post - accuracy_breakdown_v1_reports_accuracy_breakdown_post - accuracy_citation_analysis_v1_reports_accuracy_citation_analysis_post - accuracy_topic_ids_v1_reports_accuracy_topic_ids_post - accuracy_inaccurate_themes_v1_reports_accuracy_inaccurate_themes_post - accuracy_inaccurate_clusters_v1_reports_accuracy_inaccurate_clusters_post - accuracy_inaccuracy_drivers_v1_reports_accuracy_inaccuracy_drivers_post - accuracy_top_inaccurate_claims_v1_reports_accuracy_top_inaccurate_claims_post - accuracy_claim_breakdown_v1_reports_accuracy_claim_breakdown_post - accuracy_claim_citations_v1_reports_accuracy_claim_citations_post - accuracy_cluster_example_runs_v1_reports_accuracy_cluster_example_runs_post - accuracy_cluster_verification_pairs_v1_reports_accuracy_cluster_verification_pairs_post - accuracy_factcheck_setup_status_v1_reports_accuracy_factcheck_setup_status_post note: >- MCP exposes summary FactCheck scores and claims only; the 13 v1 accuracy drill-down operations have no tool. - capability: Social (YouTube) reports operations: - query_youtube_channels_v2_reports_social_youtube_channels_post - query_youtube_videos_v2_reports_social_youtube_videos_post - query_youtube_summary_v2_reports_social_youtube_summary_post note: Added to REST in August 2026; not yet surfaced as MCP tools. - capability: Content optimization operations: - optimization_list_v1_content__asset_id__optimization_get - optimization_analysis_v1_content__asset_id__optimization__content_id__get - capability: OpenAI Ads operations: - get_account_insights_v1_ads_openai_ads_ad_account_insights_get - capability: Web search results operations: - query_web_search_results_v1_reports_web_search_results_post - capability: Integrations operations: - list_integrations_v1_integrations_get - capability: Organization reference data not surfaced as tools operations: - get_assets_v1_org_assets_get - get_personas_v1_org_personas_get - get_category_assets_v1_org_categories__category_id__assets_get - get_category_personas_v1_org_categories__category_id__personas_get - get_category_citation_categories_v1_org_categories__category_id__citation_categories_get - get_category_citation_tags_v1_org_categories__category_id__citation_tags_get - create_category_prompts_v1_org_categories__category_id__prompts_post - update_category_prompts_v1_org_categories__category_id__prompts_patch - update_category_prompt_status_v1_org_categories__category_id__prompts_status_patch note: >- Prompt creation and mutation are deliberately absent from MCP, consistent with the documented read-only posture of the analytics tool set. - capability: SSE stream variants note: >- All 15 /stream operations are REST-only. MCP tools return complete result sets rather than server-sent event streams. coverage: tools_named: 30 tools_bound_to_rest: 26 mcp_only: 4 rest_operations_total: 125 rest_operations_with_a_tool: 30 rest_operations_without_a_tool: 95 note: >- Tool count is 30 across the analytics and agents capability pages. Bound operation count (30) counts distinct operations named in crosswalk[].rest; several tools bind to both a v1 and a v2 operation, and roughly three quarters of the REST surface — Shopping, Projects, Documents, accuracy drill-downs, YouTube and content optimization — has no MCP tool at all.