// SCAN REPORT

Nodebench

724 tools. 211 can modify or destroy data without limits.

9 destructive tools with no built-in limits. Policy required.

Last updated: 6 Jun 2026

211 can modify or destroy data

513 read-only

724 tools total

Free to start. No card required.

TOOL MAP

Read (513) Write / Execute (202) Destructive / Financial (9)

WHAT CAN GO WRONG

Destructive tools (compress_or_expand_text, delete_learning, dive_changelog) permanently delete resources. There is no undo. An agent calling these in a retry loop causes irreversible damage.

Write operations (add_forecast_evidence, add_rss_source, archive_content) modify state. Without rate limits, an agent can make hundreds of changes in seconds — faster than any human can review or revert.

Execute tools (benchmark_models, build_banking_packet, build_before_after_memo) trigger processes with side effects. Builds, notifications, workflows — all fired without throttling.

RECOMMENDED RULES

Deny destructive operations

{
  "compress_or_expand_text": {
    "deny_if": [
      {
        "conditions": [],
        "on_deny": "Blocked by default. Requires approval."
      }
    ]
  }
}

Destructive tools should never be available to autonomous agents without human approval.

Rate limit write operations

{
  "add_forecast_evidence": {
    "limits": [
      {
        "counter": "add_forecast_evidence_per_hour",
        "window": "hour",
        "max": 30,
        "scope": "grant"
      }
    ]
  }
}

Prevents bulk unintended modifications from agents caught in loops.

Cap read operations

{
  "abandon_cycle": {
    "limits": [
      {
        "counter": "abandon_cycle_per_minute",
        "window": "minute",
        "max": 60,
        "scope": "grant"
      }
    ]
  }
}

Controls API costs and prevents retry loops from exhausting upstream rate limits.

ALL 724 NODEBENCH TOOLS

DESTRUCTIVE 9 tools

Destructive compress_or_expand_text Destructive delete_learning Destructive dive_changelog Destructive founder_local_weekly_reset Destructive founder_weekly_reset Destructive remove_ai_signatures Destructive share_revoke_packet_link Destructive track Destructive unload_toolset

EXECUTE 90 tools

Execute benchmark_models Execute build_banking_packet Execute build_before_after_memo Execute build_causal_chain Execute build_claim_graph Execute build_company_packet Execute build_company_profile_starter Execute build_diligence_packet Execute build_founder_operating_model Execute build_investor_packet Execute build_research_digest Execute build_shared_context_subscription Execute build_shared_context_subscription_manifest Execute build_slack_onepager Execute build_submission_export Execute build_temporal_graph Execute call_driver_tool Execute call_webmcp_tool Execute compare_eval_runs Execute compile_decision_packet Execute compile_environment_spec Execute compile_scenarios Execute compile_tension_model Execute execution-trace-workflow Execute founder_direction_assessment Execute grade_agent_run Execute graphify_report Execute gtm_script_builder Execute invoke_openclaw_skill Execute invoke_view_tool Execute judge_tool_output Execute link_durable_objects Execute log_interaction Execute navigate_to_view Execute nb_start_agent Execute nodebench.research_run Execute preconditions_verified Execute primary_mission_preserved Execute record_dogfood_telemetry Execute request_execution_approval Execute retry_budget_respected Execute run_autonomous_loop Execute run_benchmark_batch Execute run_browserstack_benchmark_lane Execute run_closed_loop Execute run_code_analysis Execute run_competitor_signal_benchmark Execute run_deep_sim Execute run_dogfood_batch_with_judge Execute run_entity_intelligence_mission Execute run_flicker_detection Execute run_founder_autonomy_benchmark Execute run_graphify Execute run_judge_loop Execute run_mandatory_flywheel Execute run_oracle_comparison Execute run_packet_to_implementation_benchmark Execute run_quality_gate Execute run_recon Execute run_research_cell Execute run_self_directed_delivery_loop Execute run_self_heal Execute run_self_maintenance Execute run_signal_sweep Execute run_sync_bridge_flush Execute run_tests_cli Execute run_visual_qa_suite Execute sandbox_batch Execute sandbox_execute Execute scaffold_nodebench_project Execute scaffold_research_pipeline Execute scrapling_crawl Execute scrapling_crawl_stop Execute self_implement Execute simulate_decision_paths Execute solve_green_polygon_area_from_image Execute spawn_openclaw_agent Execute start_autonomy_benchmark Execute start_component_flow Execute start_dogfood_session Execute start_eval_run Execute start_execution_run Execute start_ui_dive Execute start_verification_cycle Execute track_milestone Execute trigger_batch_run Execute trigger_investigation Execute trigger_verify_split Execute ui-qa-checklist Execute workflowTitle

WRITE 112 tools

Write add_forecast_evidence Write add_rss_source Write archive_content Write aria_labels_complete Write assign_agent_role Write attach_execution_evidence Write bootstrap_project Write broadcast_agent_update Write complete_autonomy_benchmark Write complete_eval_run Write complete_execution_run Write complete_shared_task Write compute_ssim_analysis Write configure_channel_preferences Write connect_channels Write connect_mcp_driver Write connect_webmcp_origin Write create_forecast Write create_proof_pack Write create_task_bank Write create_visual_pr Write create_workspace_folder Write csv_select_rows Write decide_re_update Write delegate_founder_issue Write delta_memo Write disconnect_driver Write disconnect_webmcp_origin Write dive_design_issue Write dive_generate_tests Write dive_save_screenshot Write draft_email_reply Write enforce_merge_gate Write export_artifact_packet Write export_crunchbase_profile Write export_dimension_bundle Write export_pitchbook_profile Write export_pr_screenshots Write export_yc_application_context Write generate_academic_caption Write generate_countermodels Write generate_flicker_report Write generate_grid_collage Write generate_implementation_plan Write generate_parallel_agents_md Write generate_plan_delegation_packet Write generate_pr_report Write generate_proposal_memo Write generate_report Write generate_self_instructions Write generate_team_install_plan Write generate_voice_scaffold Write generate_zero_draft Write graphify_import_to_subconscious Write ingest_upload Write install_nodebench_plugin Write json_select Write manage_implementation_packets Write manage_task_list Write merge_compose_output Write merge_research_results Write nb_create_document Write nb_create_event Write nb_switch_research_tab Write nodebench.report_export_complete Write nodebench.report_export_preview Write open_core_boundary_advisor Write open_dive_dashboard Write open_local_dashboard Write open_operating_dashboard Write overstory_merge_queue Write parallel-agent-team Write plan_decompose_mission Write project-setup Write projectName Write promote_to_eval Write publish_founder_issue_packet Write publish_shared_context Write publish_to_queue Write register_component Write register_shared_context_peer Write register_skill Write reject_shared_task Write research_job_market Write resolve_forecast Write resolve_founder_issue Write resolve_gap Write retention_register_connection Write save_research_resource Write save_session_note Write send_agent_message Write send_email Write send_openclaw_message Write send_peer_message Write set_watchdog_config Write setup_local_env Write setup_operator_profile Write share_create_packet_link Write smart_select_tools Write spreadsheet-enrichment-trace Write synthesize_extension_plan Write synthesize_feature_plan Write thompson_feynman_edit Write thompson_write Write update_agents_md Write update_company_truth Write update_forecast_probability Write upsert_durable_object Write watchlist_add_entity Write workflowType Write write_workspace_file Write xlsx_select_rows

READ 513 tools

Read abandon_cycle Read accept_shared_task Read ack_shared_context Read agent-contract Read agent-delegation-with-approval-trace Read agentCount Read all_tests_green Read analyze_experiment_data Read analyze_figma_flows Read analyze_repo Read analyze_screenshot Read analyze_seo_content Read analyze_voice_config Read anomalies_logged_separately Read aria_labels_present Read ask_context Read assess_risk Read audit_openclaw_skills Read baseline_exists Read benchmark_voice_latency Read bind_local_account Read blocked_infra_classified Read bootstrap_parallel_agents Read bootstrap-parallel-agents Read burn_rate_sanity Read burst_capture Read call_llm Read call_loaded_tool Read capture_responsive_suite Read capture_surface_stats Read capture_ui_screenshot Read cheapest_valid_path Read check_agent_inbox Read check_contract_compliance Read check_design_compliance Read check_dive_drivers Read check_email_setup Read check_framework_updates Read check_git_compliance Read check_mcp_setup Read check_openclaw_setup Read check_page_performance Read check_paper_logic Read check_peer_messages Read check_plugin_update_readiness Read check_skill_freshness Read check_webmcp_setup Read check_wordpress_site Read claim_agent_task Read claim_verification_scan Read classify_failure Read claude-code-parallel Read cleanup_stale_runs Read cluster_figma_flows Read color_contrast_sufficient Read company_intelligence Read company-direction-analysis-trace Read compare Read compare_founder_directions Read compare_options Read compare_savings Read compare_workflow_paths Read competitor_brief Read compiles_clean Read component_renders Read componentName Read compute_calibration Read compute_dimension_profile Read compute_web_stability Read critter_check Read csv_aggregate Read dark_light_variants_consistent Read decision_quality_scan Read delta_brief Read delta_compare Read delta_diligence Read delta_handoff Read delta_packets Read delta_retain Read delta_review Read delta_scan Read delta_self_dogfood Read delta_watch Read Deployment Read design_voice_pipeline Read detect_company_mode Read detect_contradictions Read detect_repeated_questions Read detect_subvertical Read detect_temporal_signal Read detect_vertical Read diff_crawl Read diff_outputs Read diff_screenshots Read discover_infrastructure Read discover_tools Read discover_vision_env Read dismiss_alert Read distribution_surface_scan Read dive_auto_discover Read dive_code_locate Read dive_code_review Read dive_fix_verify Read dive_interaction_test Read dive_link_backend Read dive_preflight Read dive_record_test_step Read dive_reexplore Read dive_snapshot Read dive_walkthrough Read end_component_flow Read end_dogfood_session Read end_openclaw_session Read enrich_entity Read enrich_recon Read entity_lookup Read error_handling_present Read escalate_shared_task Read eval_scores_improved Read evidence_attached Read evidence_gap_scan Read extract_figma_frames Read extract_fractions_and_simplify_from_image Read extract_patent_claims Read extract_publication_metadata Read extract_regulatory_artifacts Read extract_structured_data Read extract_trial_evidence Read extract_variables Read extract_video_frames Read fetch_rss_feeds Read fetch_url Read fileUri Read financial_hygiene_check Read find_contradictions_for Read findTools Read flag_important_change Read focus_visible Read follows_existing_patterns Read forecast_temporal_trend Read form_labels_linked Read founder_company_naming_pack Read founder_deep_context_gather Read founder_delegation_boundary_scan Read founder_gaps_detect Read founder_local_gather Read founder_local_synthesize Read founder_materials_check Read founder_next_unlocks Read founder_ontrack_scorecard Read founder_packet_diff Read founder_packet_history_diff Read founder_packet_validate Read founder_readiness_score Read founder_stage_assess Read founder_target_customer_map Read get_ab_test_report Read get_active_forecasts Read get_agent_role Read get_annual_retrospective Read get_autopilot_status Read get_batch_run_history Read get_benchmark_history Read get_benchmark_oracles Read get_benchmark_report Read get_boilerplate_status Read get_causal_chain Read get_compaction_recovery Read get_company_truth Read get_context_bundle Read get_daily_brief_summary Read get_daily_log Read get_design_spec Read get_design_violations Read get_dimension_profile Read get_distribution_surfaces Read get_dive_report Read get_dive_tree Read get_dogfood_sessions Read get_dogfood_telemetry Read get_drift_report Read get_engine_context_health Read get_entity_graph_summary Read get_event_ledger Read get_failure_triage Read get_figma_design_context Read get_flywheel_status Read get_forecast_chain Read get_forecast_evidence Read get_forecast_track_record Read get_founder_execution_order Read get_founder_job_topology Read get_founder_packet_resource Read get_founder_progression_rubric Read get_gate_history Read get_gate_preset Read get_important_changes Read get_improvement_recommendations Read get_ingest_status Read get_judge_history Read get_latest_signals Read get_messaging_health Read get_monthly_report Read get_narrative_status Read get_observability_summary Read get_openclaw_delivery_status Read get_openclaw_results Read get_ops_dashboard Read get_packet_lineage Read get_parallel_status Read get_path_replay Read get_proactive_alerts Read get_project_context Read get_quarterly_review Read get_recon_summary Read get_regression_gate Read get_repeat_cognition_metrics Read get_role_packet_defaults Read get_self_directed_delivery_run Read get_self_eval_report Read get_sentinel_report Read get_session_journal Read get_session_profile Read get_shared_context_packet Read get_shared_context_peer Read get_shared_context_snapshot Read get_signal_recommendations Read get_source_trust_policy Read get_state_diff_history Read get_subconscious_hint Read get_sync_bridge_status Read get_system_pulse Read get_tool_graph Read get_tool_quick_ref Read get_trajectory_analysis Read get_trajectory_summary Read get_traversal_plan Read get_uptime_stats Read get_usage_insights Read get_verification_status Read get_view_capabilities Read get_view_state Read get_watchdog_log Read get_weekly_summary Read get_workflow_chain Read get_workflow_history Read getMethodology Read goal Read grade_fraction_quiz_from_image Read graphify_status Read harness_get_mission_status Read harness_list_runs Read has_regression_test Read hasHook Read hasOpinion Read hasQuestion Read heading_hierarchy Read heartbeat_shared_context_peer Read hiring_gap_scan Read important_change_review Read ingest_claude_code_sessions Read ingest_codebase_changes Read ingest_dive_screenshots Read ingest_temporal_observation Read inject_context_into_prompt Read invalidate_shared_context Read investigate Read judge_request_retry Read judge_session Read judge_verify_subtask Read keyboard_navigable Read landmark_regions_present Read learnings_banked Read learnings_documented Read list_agent_tasks Read list_available_toolsets Read list_available_views Read list_contradictions Read list_dimension_evidence Read list_dimension_interactions Read list_dimension_snapshots Read list_driver_tools Read list_eval_runs Read list_extracted_skill_templates Read list_founder_issue_packets Read list_learnings Read list_openclaw_channels Read list_pending_sync_operations Read list_self_directed_delivery_runs Read list_shared_context_peers Read list_skills Read list_stale_packets Read list_verification_cycles Read list_webmcp_tools Read list_workspace Read load_diligence_pack Read load_session_notes Read load_toolset Read loading_skeleton_present Read loading_states_handled Read log_benchmark_milestone Read log_context_budget Read log_gap Read log_phase_findings Read log_recon_finding Read log_test_result Read log_tool_call Read manipulate_screenshot Read meeting_notes_extract_decisions Read mine_session_patterns Read monitor_repo Read multi_criteria_score Read nb_engage_feed_item Read nb_filter_by_stage Read nb_get_agent_status Read nb_get_feed_items Read nb_get_funding_brief Read nb_get_leaderboard Read nb_get_pr_status Read nb_get_qa_results Read nb_get_signal_detail Read nb_get_signals Read nb_list_agents Read nb_list_deals Read nb_list_documents Read nb_list_events Read nb_list_repos Read nb_list_scenarios Read nb_list_signals Read nb_search_documents Read nb_search_research Read nb_view_screenshots Read no_console_errors Read no_critical_gaps Read no_forbidden_behaviors Read no_hardcoded_secrets Read no_layout_shift Read no_lint_warnings Read no_regressions Read no_todo_comments Read nodebench.activity_timeline Read nodebench.capture Read nodebench.expand_resource Read nodebench.notebook_append Read noGenericHashtags Read noReportHeader Read noWallOfText Read onboarding Read oracle-test-harness Read oracleSource Read orchestrating-swarms Read overstory_fleet_status Read overstory_mail_log Read overstory_qa_summary Read partnership_target_map Read pdf_search_text Read pixel_diff_within_threshold Read polish_academic_text Read pre_delegation Read predict_risks_from_patterns Read projectGoal Read projectPath Read propose_shared_task Read pull_profile Read pull_report Read pull_shared_context Read quality_gate_enforced Read query_daily_brief Read query_funding_entities Read query_graphify Read query_research_queue Read query_temporal_signals Read query_view_data Read queue_sync_operation Read rank_interventions Read rate_packet_usefulness Read read_csv_file Read read_docx_text Read read_emails Read read_image_ocr_text Read read_json_file Read read_jsonl_file Read read_pdf_text Read read_pptx_text Read read_text_file Read read_workspace_file Read read_xlsx_file Read readiness_scan Read record_eval_result Read record_event Read record_execution_decision Read record_execution_step Read record_execution_verification Read record_fix_attempt Read record_learning Read record_manual_correction Read record_path_step Read record_provenance_receipt Read record_repeated_question Read record_state_diff Read record_sync_artifact Read record_sync_outcome Read reduced_motion_respected Read refresh_subconscious Read refresh_task_context Read regression_guards_created Read release_agent_task Read render_decision_memo Read render_flow_visualization Read report Read responsive_breakpoints_intact Read responsive_check Read retention_get_status Read retention_status Read retention_sync Read retention_sync_findings Read review_paper_as_reviewer Read review_pr_checklist Read risk_assessed Read riskLevel Read route_founder_packet Read runway_check Read sandbox_ingest Read sandbox_search Read sandbox_stats Read scaffold_directory Read scaffold_openclaw_project Read scan_capabilities Read scan_dependencies Read scan_terminal_security Read scan_webmcp_origin Read scan_wordpress_updates Read score_compounding Read score_scenario_branch Read scrapling_batch_fetch Read scrapling_crawl_status Read scrapling_extract Read scrapling_fetch Read scrapling_track_element Read search Read search_all_knowledge Read search_content_archive Read search_github Read search_learnings Read seo_audit_url Read service_to_dashboard_path Read session_memory_cycle Read share_get_packet_link Read shortest_valid_path Read site_map Read sniff_record_human_review Read solve_bass_clef_age_from_image Read solve_red_green_deviation_average_from_image Read solve_storage_upgrade_cost_per_file_from_image Read storybook_story_exists Read strategicQuestion Read structured_recon Read subagentCount Read subjectCompany Read submission_readiness_score Read suggest_optimizations Read suggest_tests Read summarize Read summarize_session Read sync_company Read sync_daily_brief Read sync_figma_tokens Read sync_operator_profile Read sync_report Read sync_skill Read synthesize_integration_proposal Read synthesize_recon_to_learnings Read system_observability Read tag_ui_bug Read task Read task_success Read taskDescription Read team_alignment_check Read techStack Read tests_pass Read thompson_anti_elitism_lint Read thompson_pipeline Read thompson_quality_gate Read thompson_visual_map Read thompson-protocol Read three_layer_tests Read toon_decode Read toon_encode Read track_action Read track_entity_changes Read track_intent Read transcribe_audio_file Read translate_academic Read traverse_entity_graph Read traverse_feed Read triple_verify Read underCharLimit Read validate_agent_compatibility Read validate_shortcut Read verdict_is_defensible Read verify_concept_support Read visual_consistency Read watchlist_get_alerts Read watchlist_list_entities Read watchlist_refresh_entities Read web_search Read within_budget Read workflow_adoption_scan Read workflowGoal Read xlsx_aggregate Read zip_extract_file Read zip_list_files Read zip_read_text_file

FAQ

Can an AI agent delete data through the Nodebench MCP server? +

Yes. The Nodebench server exposes 9 destructive tools including compress_or_expand_text, delete_learning, dive_changelog. These permanently remove resources with no undo. PolicyLayer blocks destructive tools by default so they never reach the upstream server.

How do I prevent bulk modifications through Nodebench? +

The Nodebench server has 112 write tools including add_forecast_evidence, add_rss_source, archive_content. Set a rate limit in your policy -- for example, 10 calls per hour prevents an agent from making more than 10 modifications per hour. PolicyLayer enforces this at the gateway, before calls reach Nodebench.

How many tools does the Nodebench MCP server expose? +

724 tools across 4 categories: Destructive, Execute, Read, Write. 513 are read-only. 211 can modify, create, or delete data.

How do I enforce a policy on Nodebench? +

Register the Nodebench MCP server in PolicyLayer, apply the suggested rules above (adjust the limits to your use case), and point your AI client at the PolicyLayer proxy URL instead of the server directly. Your agents keep the same tools; PolicyLayer evaluates every call against policy before it executes. Nothing to install, live in minutes.

Nodebench

TOOL MAP

WHAT CAN GO WRONG

RECOMMENDED RULES

ALL 724 NODEBENCH TOOLS

FAQ

Other MCP servers with similar tools.

Enforce policy on every Nodebench tool call.

Nodebench

// TOOL MAP

// WHAT CAN GO WRONG

// RECOMMENDED RULES

// ALL 724 NODEBENCH TOOLS

// FAQ

Other MCP servers with similar tools.

Enforce policy on every Nodebench tool call.

TOOL MAP

WHAT CAN GO WRONG

RECOMMENDED RULES

ALL 724 NODEBENCH TOOLS

FAQ