{"generated_at":"2026-10-14T20:07:22+00:00","item":{"added_at":"2026-07-30","analysis_method":"independent_manual_review_2026-10-10","assessment_status":"provisional","auth_posture":"auth_required","category":"Web Connector","confidence":"medium","description":"elluminate connects Claude to AI evaluation and observation workflows. Create evaluation collections, criteria, prompt templates, model configurations, and experiments. Upload agent results and inspect experiment status, failed examples, annotations, and trace steps.","did":"4fc2fa9c-bf2b-4d29-ad27-b3b9681f5b30","documentation_url":"https://docs.elluminate.de/","editorial_description":{"extended_description":"Tools organize evaluations into projects and organizations, clone existing configurations, maintain template variables, and compare response samples. Experiment completion can be monitored before reviewing detailed responses.","schema_version":1,"summary":"elluminate connects Claude to AI evaluation and observation workflows.\nCreate evaluation collections, criteria, prompt templates, model configurations, and experiments.\nUpload agent results and inspect experiment status, failed examples, annotations, and trace steps."},"homepage_url":"","icon":"","id":"4fc2fa9c-bf2b-4d29-ad27-b3b9681f5b30","installs":0,"last_scanned":"","license":"","long_description":"Tools organize evaluations into projects and organizations, clone existing configurations, maintain template variables, and compare response samples. Experiment completion can be monitored before reviewing detailed responses.","manual_assessment":"high","manual_review":{"all_findings":[{"evidence":"Tool set creates experiments and waits on their execution.","reasoning":"Starting model-evaluation workloads is a High operational action.","risk_type":"deployment_infrastructure","severity":"high","title":"Can create and run experiments","triggering_tools_or_skills":["create_experiment","wait_for_experiment_completion","get_experiment_status"]},{"evidence":"Names explicitly delete criteria/template variables.","reasoning":"Deleting meaningful experiment configuration is High.","risk_type":"data_deletion","severity":"high","title":"Can delete evaluation configuration","triggering_tools_or_skills":["delete_criterion","delete_template_variable"]},{"evidence":"Inventory declares persistent configuration and result uploads.","reasoning":"These modify workspace data.","risk_type":"data_modification","severity":"medium","title":"Can create and upload evaluation assets","triggering_tools_or_skills":["create_project","create_llm_config","create_prompt_template","upload_agent_results"]}],"assessment_status":"provisional","auth_and_scopes":"Authenticated elluminate organization/project access is implied; scopes and compute billing are undocumented.","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/4fc2fa9c-bf2b-4d29-ad27-b3b9681f5b30.json","what_it_proves":"Candidate PR #125 record identity (name \"elluminate\", publisher, type web_connector) and 45 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/4fc2fa9c-bf2b-4d29-ad27-b3b9681f5b30: complete tool_names inventory (45 tools), remote streamable-http endpoint https://app.elluminate.de/mcp, auth_posture auth_required, directory tier community"}],"primary_reason":"elluminate can initiate experiments, delete evaluation configuration, and upload results.","provisional_risk":null,"publication_recommendation":"accept","requires_runtime_verification":true,"review_key":"claude:4fc2fa9c-bf2b-4d29-ad27-b3b9681f5b30","reviewed_at":"2026-10-10","reviewer_notes":"All 45 names reviewed; experiment resource and billing controls require runtime verification.","risk_domains":{"arbitrary_network_access":"false","authz_ambiguity":"unknown","browser_control":"false","credential_or_admin_actions":"false","deletes_data":"true","deployment_or_infrastructure":"true","dynamic_hidden_actions":"false","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":"false","spends_or_moves_money":"false"},"tools_or_skills_reviewed":45,"tools_or_skills_total":45},"name":"elluminate","needs_runtime_verification":true,"plutonium_url":"https://plutonium.pluto.security/detail.html?planet=claudesec&did=4fc2fa9c-bf2b-4d29-ad27-b3b9681f5b30&utm_source=plutonium_analysis_skill&utm_medium=claude_skill&utm_campaign=connector_risk_assessment","privacy_policy_url":"https://app.elluminate.de/privacy-policy","provisional_risk":null,"publication_recommendation":"accept","publisher":"ellamind","publisher_url":"https://ellamind.com/","remote_url":"https://app.elluminate.de/mcp","repository_url":"","risk":"high","risk_severity":"high","security_risks":[{"description":"This connector can deploy code or change infrastructure and production environments. Review findings: Starting model-evaluation workloads is a High operational action.","evidence":"Can create and run experiments: Tool set creates experiments and waits on their execution.","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":"Can create and run experiments","triggering_tools":["create_experiment","wait_for_experiment_completion","get_experiment_status"]},{"description":"This connector can permanently remove data or resources. Review findings: Deleting meaningful experiment configuration is High.","evidence":"Can delete evaluation configuration: Names explicitly delete criteria/template variables.","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":"Can delete evaluation configuration","triggering_tools":["delete_criterion","delete_template_variable"]},{"description":"This connector can create, update, configure, or otherwise change connected resources. Review findings: These modify workspace data.","evidence":"Can create and upload evaluation assets: Inventory declares persistent configuration and result uploads.","manual_risk_type":"data_modification","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":"medium","title":"Can create and upload evaluation assets","triggering_tools":["create_project","create_llm_config","create_prompt_template","upload_agent_results"]}],"signature_status":"unknown","source_code_reviewed":false,"support_url":"help@elluminate.de","tags":["code_deployment","deletes_data","modifies_data","network_access","oauth_required","web_connector"],"tools":[{"description":"","name":"add_criteria_to_criterion_set"},{"description":"","name":"add_criterion_set_to_prompt_template"},{"description":"","name":"add_template_variables_to_collection"},{"description":"","name":"annotate_response"},{"description":"","name":"clone_collection"},{"description":"","name":"clone_criterion_set"},{"description":"","name":"clone_llm_config"},{"description":"","name":"clone_project"},{"description":"","name":"clone_prompt_template"},{"description":"","name":"create_collection"},{"description":"","name":"create_criterion_set"},{"description":"","name":"create_experiment","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: Starting model-evaluation workloads is a High operational action."}},{"description":"","name":"create_llm_config","risk":{"category":"state_change","level":"medium","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: These modify workspace data."}},{"description":"","name":"create_project","risk":{"category":"state_change","level":"medium","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: These modify workspace data."}},{"description":"","name":"create_prompt_template","risk":{"category":"state_change","level":"medium","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: These modify workspace data."}},{"description":"","name":"delete_criterion","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: Deleting meaningful experiment configuration is High."}},{"description":"","name":"delete_template_variable","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: Deleting meaningful experiment configuration is High."}},{"description":"","name":"edit_criterion"},{"description":"","name":"get_agent_results_definition"},{"description":"","name":"get_annotated_responses"},{"description":"","name":"get_collection"},{"description":"","name":"get_criterion_set"},{"description":"","name":"get_experiment"},{"description":"","name":"get_experiment_map"},{"description":"","name":"get_experiment_status","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: Starting model-evaluation workloads is a High operational action."}},{"description":"","name":"get_failed_examples"},{"description":"","name":"get_llm_config"},{"description":"","name":"get_project"},{"description":"","name":"get_prompt_template"},{"description":"","name":"get_response_by_id"},{"description":"","name":"get_response_details"},{"description":"","name":"get_template_variables"},{"description":"","name":"get_trace_step"},{"description":"","name":"get_trace_steps"},{"description":"","name":"list_collections"},{"description":"","name":"list_comparison_samples"},{"description":"","name":"list_criterion_sets"},{"description":"","name":"list_experiments"},{"description":"","name":"list_llm_configs"},{"description":"","name":"list_organizations"},{"description":"","name":"list_projects"},{"description":"","name":"list_prompt_templates"},{"description":"","name":"update_template_variable"},{"description":"","name":"upload_agent_results","risk":{"category":"state_change","level":"medium","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: These modify workspace data."}},{"description":"","name":"wait_for_experiment_completion","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: Starting model-evaluation workloads is a High operational action."}}],"tools_count":45,"transport":"streamable-http","type":"web_connector","url":"https://claude.ai/directory/4fc2fa9c-bf2b-4d29-ad27-b3b9681f5b30","uuid":"eb05b475-3372-5153-90cf-7029cf5f87cb","verification_tier":"community","version":""},"query_key":"elluminate","schema_version":1,"status":"match"}