{"generated_at":"2026-10-14T20:07:22+00:00","item":{"added_at":"2026-08-14","analysis_method":"independent_manual_review_2026-10-10","assessment_status":"provisional","auth_posture":"auth_required","category":"Web Connector","confidence":"medium","description":"Ballet builds and runs AI agents and multi-step automation playbooks. Manage agents, skills, tools, and MCP connections, then execute workflows. Inspect runs and individual steps through authenticated account access.","did":"8df24de5-5d00-47aa-82cb-9845f0219e9b","documentation_url":"https://docs.ballet.dev/","editorial_description":{"extended_description":"The platform supports scheduled and event-driven workflows, including webhooks, Slack commands, and forms. Connector tools can create, update, or delete automation components and connected-server configurations.","schema_version":1,"summary":"Ballet builds and runs AI agents and multi-step automation playbooks.\nManage agents, skills, tools, and MCP connections, then execute workflows.\nInspect runs and individual steps through authenticated account access."},"homepage_url":"","icon":"","id":"8df24de5-5d00-47aa-82cb-9845f0219e9b","installs":0,"last_scanned":"","license":"","long_description":"The platform supports scheduled and event-driven workflows, including webhooks, Slack commands, and forms. Connector tools can create, update, or delete automation components and connected-server configurations.","manual_assessment":"high","manual_review":{"all_findings":[{"evidence":"These tools remove durable automation resources.","reasoning":"They can destroy production workflow logic and integrations.","risk_type":"data_deletion","severity":"high","title":"Deletes agents, servers, playbooks, skills and tools","triggering_tools_or_skills":["delete_agent","delete_mcp_server","delete_playbook","delete_skill","delete_tool"]},{"evidence":"The listing says playbooks orchestrate multiple agents and external actions.","reasoning":"Execution can invoke arbitrary configured tools and side effects beyond the fixed inventory.","risk_type":"code_execution","severity":"high","title":"Runs multi-agent playbooks with connected tools","triggering_tools_or_skills":["run_playbook"]},{"evidence":"These tools configure continuously triggered operational workflows.","reasoning":"They materially change automation infrastructure.","risk_type":"deployment_infrastructure","severity":"high","title":"Creates/updates persistent automation and connected MCP servers","triggering_tools_or_skills":["create_mcp_server","connect_mcp_server","create_agent","create_playbook"]},{"evidence":"These actions configure privileged integrations and their capabilities.","reasoning":"They are integration/credential administration.","risk_type":"credential_admin","severity":"high","title":"Connects and manages external MCP servers/tools","triggering_tools_or_skills":["connect_mcp_server","create_mcp_server","update_mcp_server"]}],"assessment_status":"provisional","auth_and_scopes":"Authentication is required; agents/playbooks/tools/MCP-server configuration and runs follow Ballet workspace roles, which are not documented.","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/8df24de5-5d00-47aa-82cb-9845f0219e9b.json","what_it_proves":"Candidate PR #125 record identity (name \"Ballet\", publisher, type web_connector) and 31 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/8df24de5-5d00-47aa-82cb-9845f0219e9b: complete tool_names inventory (31 tools), remote streamable-http endpoint https://api.ballet.dev/mcp, auth_posture auth_required, directory tier community"}],"primary_reason":"Ballet can execute opaque multi-tool playbooks, destroy automation resources and administer integrations.","provisional_risk":null,"publication_recommendation":"accept","requires_runtime_verification":true,"review_key":"claude:8df24de5-5d00-47aa-82cb-9845f0219e9b","reviewed_at":"2026-10-10","reviewer_notes":"Reviewed all 31 tools. Runtime playbook/tool contents are dynamic; exact downstream effects require per-workspace inspection but multiple fixed High controls are explicit.","risk_domains":{"arbitrary_network_access":"false","authz_ambiguity":"true","browser_control":"false","credential_or_admin_actions":"true","deletes_data":"true","deployment_or_infrastructure":"true","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":"true","sends_messages_as_user":"false","spends_or_moves_money":"false"},"tools_or_skills_reviewed":31,"tools_or_skills_total":31},"name":"Ballet","needs_runtime_verification":true,"plutonium_url":"https://plutonium.pluto.security/detail.html?planet=claudesec&did=8df24de5-5d00-47aa-82cb-9845f0219e9b&utm_source=plutonium_analysis_skill&utm_medium=claude_skill&utm_campaign=connector_risk_assessment","privacy_policy_url":"https://www.brainfishai.com/privacy","provisional_risk":null,"publication_recommendation":"accept","publisher":"Brainfish","publisher_url":"https://ballet.dev/","remote_url":"https://api.ballet.dev/mcp","repository_url":"","risk":"high","risk_severity":"high","security_risks":[{"description":"This connector can permanently remove data or resources. Review findings: They can destroy production workflow logic and integrations.","evidence":"Deletes agents, servers, playbooks, skills and tools: These tools remove durable automation resources.","manual_risk_type":"data_deletion","remediation":{"block_tool_categories":["destructive"],"steps":["Disable destructive tools unless deletion is an explicit workflow requirement.","Require human confirmation and verified backups before deletion."]},"risk_type":"deletes_data","severity":"high","title":"Deletes agents, servers, playbooks, skills and tools","triggering_tools":["delete_agent","delete_mcp_server","delete_playbook","delete_skill","delete_tool"]},{"description":"This connector can execute code, commands, queries, or dynamically selected actions. Review findings: Execution can invoke arbitrary configured tools and side effects beyond the fixed inventory.","evidence":"Runs multi-agent playbooks with connected tools: The listing says playbooks orchestrate multiple agents and external actions.","manual_risk_type":"code_execution","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":"Runs multi-agent playbooks with connected tools","triggering_tools":["run_playbook"]},{"description":"This connector can deploy code or change infrastructure and production environments. Review findings: They materially change automation infrastructure.","evidence":"Creates/updates persistent automation and connected MCP servers: These tools configure continuously triggered operational workflows.","manual_risk_type":"deployment_infrastructure","remediation":{"block_tool_categories":["code_execution","state_change"],"steps":["Default to a sandbox or staging environment.","Require code review, human approval, audit logging, and a tested rollback path for production changes."]},"risk_type":"code_deployment","severity":"high","title":"Creates/updates persistent automation and connected MCP servers","triggering_tools":["create_mcp_server","connect_mcp_server","create_agent","create_playbook"]},{"description":"This connector can create, update, configure, or otherwise change connected resources. Review findings: They are integration/credential administration.","evidence":"Connects and manages external MCP servers/tools: These actions configure privileged integrations and their capabilities.","manual_risk_type":"credential_admin","remediation":{"block_tool_categories":["state_change"],"steps":["Use a test account or sandbox before allowing changes to production data.","Grant only the write permissions required for the intended workflow.","Require human confirmation for consequential state changes."]},"risk_type":"modifies_data","severity":"high","title":"Connects and manages external MCP servers/tools","triggering_tools":["connect_mcp_server","create_mcp_server","update_mcp_server"]}],"signature_status":"unknown","source_code_reviewed":false,"support_url":"https://docs.ballet.dev","tags":["code_deployment","deletes_data","modifies_data","network_access","oauth_required","runs_code","web_connector"],"tools":[{"description":"","name":"connect_mcp_server","risk":{"category":"state_change","level":"high","recommendation":"Follow the connector-level remediation and require approval for consequential actions.","why":"This connector can create, update, configure, or otherwise change connected resources. Review findings: They are integration/credential administration."}},{"description":"","name":"create_agent","risk":{"category":"code_execution","level":"high","recommendation":"Follow the connector-level remediation and require approval for consequential actions.","why":"This connector can deploy code or change infrastructure and production environments. Review findings: They materially change automation infrastructure."}},{"description":"","name":"create_mcp_server","risk":{"category":"state_change","level":"high","recommendation":"Follow the connector-level remediation and require approval for consequential actions.","why":"This connector can create, update, configure, or otherwise change connected resources. Review findings: They are integration/credential administration."}},{"description":"","name":"create_playbook","risk":{"category":"code_execution","level":"high","recommendation":"Follow the connector-level remediation and require approval for consequential actions.","why":"This connector can deploy code or change infrastructure and production environments. Review findings: They materially change automation infrastructure."}},{"description":"","name":"create_skill"},{"description":"","name":"create_tool"},{"description":"","name":"delete_agent","risk":{"category":"destructive","level":"high","recommendation":"Follow the connector-level remediation and require approval for consequential actions.","why":"This connector can permanently remove data or resources. Review findings: They can destroy production workflow logic and integrations."}},{"description":"","name":"delete_mcp_server","risk":{"category":"destructive","level":"high","recommendation":"Follow the connector-level remediation and require approval for consequential actions.","why":"This connector can permanently remove data or resources. Review findings: They can destroy production workflow logic and integrations."}},{"description":"","name":"delete_playbook","risk":{"category":"destructive","level":"high","recommendation":"Follow the connector-level remediation and require approval for consequential actions.","why":"This connector can permanently remove data or resources. Review findings: They can destroy production workflow logic and integrations."}},{"description":"","name":"delete_skill","risk":{"category":"destructive","level":"high","recommendation":"Follow the connector-level remediation and require approval for consequential actions.","why":"This connector can permanently remove data or resources. Review findings: They can destroy production workflow logic and integrations."}},{"description":"","name":"delete_tool","risk":{"category":"destructive","level":"high","recommendation":"Follow the connector-level remediation and require approval for consequential actions.","why":"This connector can permanently remove data or resources. Review findings: They can destroy production workflow logic and integrations."}},{"description":"","name":"get_agent"},{"description":"","name":"get_mcp_server"},{"description":"","name":"get_playbook"},{"description":"","name":"get_run"},{"description":"","name":"get_run_steps"},{"description":"","name":"get_skill"},{"description":"","name":"get_tool"},{"description":"","name":"list_agents"},{"description":"","name":"list_mcp_servers"},{"description":"","name":"list_mcp_tools"},{"description":"","name":"list_playbooks"},{"description":"","name":"list_runs"},{"description":"","name":"list_skills"},{"description":"","name":"list_tools"},{"description":"","name":"run_playbook","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: Execution can invoke arbitrary configured tools and side effects beyond the fixed inventory."}},{"description":"","name":"update_agent"},{"description":"","name":"update_mcp_server","risk":{"category":"state_change","level":"high","recommendation":"Follow the connector-level remediation and require approval for consequential actions.","why":"This connector can create, update, configure, or otherwise change connected resources. Review findings: They are integration/credential administration."}},{"description":"","name":"update_playbook"},{"description":"","name":"update_skill"},{"description":"","name":"update_tool"}],"tools_count":31,"transport":"streamable-http","type":"web_connector","url":"https://claude.ai/directory/8df24de5-5d00-47aa-82cb-9845f0219e9b","uuid":"88d893d6-074a-5a51-9689-1df4cf0bd7b9","verification_tier":"community","version":""},"query_key":"ballet","schema_version":1,"status":"match"}