Pentest Swarm AI is an open-source autonomous penetration testing framework built around a true swarm architecture. Unlike pipeline-based tools that dispatch agents in a fixed order, Pentest Swarm AI uses a stigmergic blackboard: agents coordinate by reading and writing findings to shared state, allowing attack chains to emerge naturally without a central planner. One command, one API key, and the swarm handles the rest.Documentation Index
Fetch the complete documentation index at: https://mintlify.com/Armur-Ai/Pentest-Swarm-AI/llms.txt
Use this file to discover all available pages before exploring further.
Quickstart
Run your first autonomous pentest scan in under 5 minutes with a single command.
Installation
Install via Homebrew, Docker, or Go — pick the method that fits your workflow.
CLI Reference
Every command, subcommand, and flag documented with real examples and defaults.
Swarm Architecture
Understand how the stigmergic blackboard and pheromone system coordinate agents.
What makes this a swarm?
Most “multi-agent” security tools are pipelines: a planner dispatches to specialists in a fixed sequence. Pentest Swarm AI is built differently:Stigmergy
Agents coordinate through a shared blackboard, not a central orchestrator. Findings carry pheromone weights that guide other agents toward high-value targets.
Emergence
Attack chains appear that no single agent planned. A recon finding triggers the classifier; a high-severity classification wakes the exploit agent automatically.
Decentralization
Each agent runs its own trigger predicate. Add a new agent with its own predicate and it joins the swarm without rewriting the orchestrator.
Get started in three steps
Install pentestswarm
Install via Homebrew on macOS, pull the Docker image, or use
go install for the latest build.Set your API key
Run
pentestswarm init to store your Claude API key in the OS keychain, or export it as an environment variable.Explore the documentation
Bug Bounty Guide
Import scope from HackerOne, Bugcrowd, or Intigriti and generate platform-ready submission reports.
GitHub Actions
Run automated security scans in CI with SARIF output that lands in the GitHub Security tab.
Playbooks
Use and author YAML playbooks for OWASP Top 10, bug bounty, ASM, and CTF scenarios.
MCP Integration
Connect pentestswarm as an MCP server to Claude Desktop or Cursor for AI-assisted pentesting.
Security Hardening
Four-layer defense against memory-injection attacks: pheromone clamping, Ed25519 provenance, and more.
Deployment
Deploy with Docker Compose, configure Postgres and Redis, and run in production environments.