Falcon MCP
The MCP server connects agents to CrowdStrike Falcon for detection analysis, host information and security actions.
- 281 ★ on the repository
- Last commit
- MIT
Broad access
Why this level
- Enabled mutating tools let an agent change protection policies, exclusions and host groups under the API client’s permissions.
Install
- Works in
Manual install
uv tool install falcon-mcpFirst recommended README installer. Configure Falcon credentials and restrict tools afterward.
Access and checks
How we reviewBroad accessThe tool can change data, run commands or handle secrets.
- Enabled mutating tools let an agent change protection policies, exclusions and host groups under the API client’s permissions.
This is third-party code. Review the repository files before installing.
What it does
Falcon MCP exposes CrowdStrike Falcon data and operations through selected modules. An analyst can investigate detections, identify associated hosts and correlate findings with threat intelligence. Investigation servers can be restricted to read-only tools. Broader permissions expose policy, host-group, exclusion and other protection changes. Administrators select modules and individual tools at startup. Dynamic mode lets an agent discover a capability before loading its schema and executing it.
Who it is for. Security operations analysts investigating CrowdStrike Falcon events while controlling the actions exposed to an agent.
Good fit when
- After a detection, an analyst needs related hosts and threat context.
- For an investigation, the team needs a server without configuration-changing tools.
- In a large deployment, the agent should load only the required tool schemas.
- Before granting access, an administrator wants to inspect the actions available to the agent.
Not a fit when
- The team lacks the required Falcon API access.
- The server must be network-accessible without request authentication.
- Policy changes are forbidden but the configuration still exposes mutating tools.
Example request
In read-only mode, collect the selected detection and related host details, then list missing investigation evidence.MCP
- Transport
- stdio, http
- Authentication
- API key
Environment variables
FALCON_CLIENT_IDrequired- Falcon API client identifier
FALCON_CLIENT_SECRETrequired, secret- Falcon API client secret
FALCON_BASE_URLrequired- API URL for the required Falcon cloud
Limitations
Available data depends on the Falcon API client’s permissions and platform capabilities. HTTP access has no incoming-request authentication by default, so wider network exposure needs explicit protection. The `--tools` option adds tools to enabled modules; use `--read-only` to exclude mutations. The project is pre-1.0, and tool names, parameters and response shapes can change between minor releases, so pin versions and review changes before upgrades.
How to disable. Stop Falcon MCP and remove its AI-client connection. Revoke API credentials if access is no longer needed; undo completed Falcon changes separately.
FAQ
How do I exclude configuration changes?
Start with `--read-only`; mutating tools will not be available to the agent.
Can I expose only two tools?
Yes. List them through `--tools` without enabling modules.
What happens if a tool name is misspelled?
Startup fails so an incorrect restriction cannot silently go unnoticed.
Where can I see the actual available operations?
Request the enabled-tool list; startup logs also show active restrictions.
Why is a tool unavailable in dynamic mode?
The same restrictions apply. The response identifies the rule withholding the action.
Discussion
No questions or reviews yet.
Related
MCP serversSecurity Detections MCP: a detection rule database
Security Detections MCP
An MCP server that merges Sigma, Splunk, Elastic and KQL rules into one database for finding detections and ATT&CK gaps
An offensive security MCP server: the agent runs 150+ pentest and recon tools and drives automated target assessments
MCP platform for incident response: AI analyzes forensic artifacts and a human approves every finding with a signature.
Security scanner for MCP servers that looks for hardcoded keys, tool poisoning and injection flaws in repository code.
MCP Observatory: security testing for MCP servers
MCP Observatory
An MCP server scanner: checks schemas, drift and attack behavior, runs in CI and as an MCP server for agents
An MCP server that gives the agent vulnerability triage: NVD data, EPSS, the CISA KEV catalog, MITRE ATT&CK mapping and one-call CVE risk scoring
README in short
CrowdStrike maintains the repository with the community; the README distinguishes it from an official company product. Container deployment is available.