Crowdsentinel MCP server
io.github.thomasxm/crowdsentinel-mcp-server·v0.6.0·206 starsAI threat hunting & incident response for Elasticsearch/OpenSearch with endpoint & network forensics
206 stars67 downloads/wk
Reviews
Write oneNobody has reviewed Crowdsentinel yet.
If you have run it, two minutes of your experience saves the next person an afternoon.
Crowdsentinel tools (100, 24 write)
write = sends, deletes, buys or postsRead from the package source without running it. The installed server may list more.
add_iocs_to_investigationManually add IoCs to an investigation.
analyze_failed_loginsAnalyze failed login attempts to detect potential brute force attacks. This tool aggregates failed login events (Event ID 4625) and identifies suspicious patterns.
analyze_kill_chain_stagewrite actionanalyze_process_creationAnalyze process creation events (Event ID 4688) for suspicious activity. This is crucial for detecting malicious process execution and LOLBins usage.
analyze_search_resultsAnalyze search results and provide intelligent insights with follow-up recommendations. This is a critical tool for incident response - it acts like an experienced analyst by extracting IoCs, mapping to MITRE ATT&CK, assessing severity, and recommending next steps.
build_unified_timelineBuild a unified chronological timeline merging events from all sources.
check_esql_supportCheck if the connected Elasticsearch cluster supports ES|QL.
cleanup_storageCleanup investigation storage and enforce size limits.
close_investigationClose an investigation with a resolution summary.
correlate_by_timeFind co-occurring SIEM + endpoint activity in a time window.
correlate_siem_with_endpointCross-correlate SIEM IoCs with live endpoint data via Velociraptor.
create_data_streamwrite actionCreate a new data stream.
create_indexwrite actionCreate a new index.
create_investigationwrite actionCreate a new investigation and make it active.
decode_commandDecode an obfuscated/encoded command line and recover the IoCs + ATT&CK techniques it conceals. Encoded payloads (e.g. `powershell -EncodedCommand <base64>`) hide their C2 IP/URL from plaintext hunting and from analyze_search_results — this tool peels the encoding (stdlib base64/UTF-16LE/gzip/hex; optional `chepy` for the long tail) so the hidden indicators become visible.
delete_aliaswrite actionDelete an alias for a specific index.
delete_by_querywrite actionDeletes documents matching the provided query.
delete_data_streamwrite actionDelete one or more data streams.
delete_documentwrite actionDelete a document by ID.
delete_indexwrite actionDelete an index.
detect_schema_for_indexDetect which schema to use for an index pattern.
discover_all_assetsDiscover all assets in the Elasticsearch cluster including indices, data streams, and their metadata. This will scan all indices and extract metadata like OS type, log source, field mappings, etc.
discover_esql_indicesDiscover available indices for ES|QL hunting.
endpoint_hunt_from_iocPivot an Elastic IoC into a targeted Velociraptor hunt (Elastic -> endpoint).
endpoint_to_siem_pivotCollect an endpoint artifact, extract IoCs, then search SIEM for those IoCs across all hosts.
enrich_iocsEnrich IoCs from the active investigation with external threat intelligence.
eql_deletewrite actionDelete an async EQL search by ID.
eql_get_statusGet the status of an async EQL search.
eql_searchwrite actionExecute an EQL (Event Query Language) query for advanced threat hunting. EQL is designed for event-based data and is excellent for detecting attack patterns and sequences.
esql_querywrite actionExecute a raw ES|QL query for ad-hoc threat hunting.
execute_detection_rulewrite actionExecute a specific detection rule against Elasticsearch.
execute_esql_huntwrite actionExecute a curated ES|QL hunting query from the rule library.
execute_multiple_ruleswrite actionExecute multiple detection rules in batch.
export_iocsExport IoCs from an investigation in various formats.
export_to_mispExport investigation IoCs as a MISP event.
general_api_requestPerform a general HTTP API request. Use this tool for any Elasticsearch/OpenSearch API that does not have a dedicated tool.
generate_investigation_reportGenerate a comprehensive investigation report from multiple analyses. This tool aggregates results from multiple queries and analyses to create a complete incident response report.
get_aliasGet alias information for a specific index.
get_cluster_healthReturns basic information about the health of the cluster.
get_cluster_statsReturns high-level overview of cluster statistics.
get_data_streamGet information about one or more data streams.
get_diamond_model_guideGet the Diamond Model of Intrusion Analysis guide.
get_documentGet a document by ID.
get_enrichment_statusCheck which threat intelligence providers are configured.
get_esql_execution_historyGet ES|QL query execution history for token usage analysis.
get_esql_hunt_detailsGet detailed information about a specific ES|QL hunting rule.
get_event_type_fieldsGet all fields available for a specific event type.
get_field_mappingGet the actual field name for a semantic field concept.
get_host_activity_timelineGet a complete timeline of all activity for a specific host. This is essential for forensic analysis and incident investigation.
get_indexReturns information (mappings, settings, aliases) about one or more indices.
get_index_metadataGet metadata for a specific index pattern.
get_indices_by_typeGet indices matching a specific log type.
get_investigation_queryGet the Elasticsearch query for an investigation prompt.
get_investigation_summaryGet a comprehensive summary of an investigation.
get_kill_chain_overviewwrite actionGet a complete overview of the Cyber Kill Chain framework.
get_progressive_disclosureGet progressive disclosure prompt for session start.
get_pyramid_of_pain_guideGet the Pyramid of Pain framework guide for IoC prioritization.
get_rule_detailsGet detailed information about a specific detection rule.
get_rule_statisticsGet comprehensive statistics about the detection rule library.
get_saved_assetsRetrieve previously discovered assets from the saved file.
get_schema_detailsGet detailed information about a specific schema.
get_shared_iocsGet IoCs shared across investigations and tools.
get_system_statusReport local CrowdSentinel system status and connector readiness.
hunt_adjacent_stageshunt_by_kill_chain_stagewrite actionhunt_by_timeframeHunt for specific attack patterns within a timeframe. This tool automatically searches for common attack indicators based on predefined patterns.
hunt_for_iocHunt for a specific Indicator of Compromise (IoC) across logs. This tool searches multiple fields based on the IoC type.
hunt_suspicious_process_activityhunt_with_rule_categorywrite actionExecute all rules in a specific category for comprehensive hunting.
hunt_with_sigma_rulesHunt for threats in EVTX logs using Sigma rules.
index_documentwrite actionCreates or updates a document in the index.
investigate_with_promptwrite actionExecute a specific investigation prompt against Elasticsearch.
iterative_huntlist_aliasesList all aliases.
list_available_capabilitiesList available CrowdSentinel capabilities grouped by category.
list_available_schemasList all available log source schemas in the registry.
list_detection_rulesList available detection rules with optional filtering.
list_esql_huntsList available ES|QL hunting queries from the curated rule library.
list_indicesList all indices.
list_investigationsList recent investigations with optional filtering.
lookup_iocLook up a single IoC across all configured threat intelligence providers.
map_events_to_kill_chainwrite actionMap Elasticsearch events to Cyber Kill Chain stages.
osquery_processesList running processes via osquery (current host state).
osquery_querywrite actionRun an ad-hoc READ-ONLY osquery SQL query.
osquery_socketsList process network sockets via osquery (active connections + listeners).
osquery_usersList user accounts via osquery. Captures username IoCs.
put_aliaswrite actionCreate or update an alias for a specific index.
quick_countFast triage aggregation - counts without returning documents.
quick_triageQuick triage: Run top N priority prompts for fast initial assessment.
resume_investigationResume an existing investigation and make it active.
search_documentsSearch for documents.
search_ioc_in_evtxsearch_rules_by_mitre_attackSearch detection rules by MITRE ATT&CK tactic.
search_with_lucenewrite actionExecute a Lucene query string search for flexible threat hunting. Use this for custom queries that don't fit predefined patterns.
show_investigation_promptsDisplay investigation triage questions for SIEM/SOAR.
smart_searchToken-efficient search with automatic summarization and PAGINATION.
start_guided_investigationStart a guided investigation with all triage prompts for a platform.
suggest_rules_for_dataAnalyze your data and suggest applicable detection rules.
threat_hunt_searchvalidate_rule_for_dataValidate if a detection rule is applicable to your data.
Public scan report
scanner v0.1.9 · 2026-09-25 · same rubric, same numbers if you re-run it
- Code scan123 source files scanned25/25
- –Live reliabilityno gateway calls yet and no remote to proben/a
- –Tool poisoningtools not inspected (local package is not executed); not countedn/a
- Auth qualitystatic API keys via environment variables6/15
- Maintenancelast push 68 days ago12/15
- Maintainer identityregistry namespace matches repository owner; GitHub account older than a year8/10
How Crowdsentinel compares
Against every graded MCP server, from our own scans and uptime checks. Updated after each scan run.
- Safety score
- 78/100
- higher than 57% of 48,273 servers
- Tools
- 100
- the typical server has 7
- Tools that change things
- 24%
- typical: 0%. Fewer write tools means less an agent can break
- GitHub stars
- 206
- more than 95% of 28,534 servers
- Downloads a week
- 67
- more than 54% of 11,915 servers with a package
What its tools let an agent do: 72 read · 5 create · 14 change · 5 delete · 4 run or deploy
Install Crowdsentinel in Claude Code, Cursor or VS Code
claude mcp add crowdsentinel-mcp-server -- uvx crowdsentinel-mcp-server
What the publisher says
From the Crowdsentinel repository's README, as published. We do not edit it. Read it on GitHub
<!-- mcp-name: io.github.thomasxm/crowdsentinel-mcp-server -->
CrowdSentinel MCP Server
AI-Powered Threat Hunting & Incident Response Framework
MCP Official Registry · PyPI Package
Open-source threat hunting orchestrator connecting LLMs to enterprise security data via Model Context Protocol (MCP)
Quick Start · Installation · CLI Usage · Features · Architecture · Documentation · Examples
Warning
This project is in active development and intended for security testing, research, and educational purposes only. It is not production-ready. Do not deploy in production environments. APIs, tool interfaces, and data formats may change without notice. Use at your own risk.
Demo
https://github.com/user-attachments/assets/0d0381f0-5b68-43b2-8630-19ec130885b2
What is CrowdSentinel?
CrowdSentinel transforms traditional SIEM querying into intelligent, framework-driven investigations using natural language. It serves as a unified security intelligence layer that connects large language models to enterprise security data sources, enabling:
Shortened. The full README is on GitHub.
Nothing above is checked by us. What we check is on the safety report.
Crowdsentinel: common questions
- Is Crowdsentinel MCP server safe?
- Mostly: it is graded B (78/100). Read the Crowdsentinel safety report
- How do I install Crowdsentinel?
- It runs on your machine. Copy the Claude Code, Cursor, VS Code or Claude Desktop config from the install section.
- Does Crowdsentinel need an API key?
- Yes. The registry entry asks for
ELASTICSEARCH_API_KEY,ELASTICSEARCH_PASSWORD,ELASTICSEARCH_BEARER_TOKEN. - Is Crowdsentinel maintained?
- The last commit was 106 days ago (2026-07-19). The latest release is v0.6.0.
- How does Crowdsentinel compare with other MCP servers?
- safety score 78/100 (higher than 57% of 48,273 servers); tools 100 (the typical server has 7); tools that change things 24% (typical: 0%. Fewer write tools means less an agent can break). Full comparison.