# Armur-Ai/Pentest-Swarm-AI

Autonomous penetration testing using a swarm of AI agents. Orchestrates recon, classification, exploitation, and reporting specialists with ReAct reasoning — supports bug bounty, continuous monitoring, and CTF modes. Built with Go, Claude API, and 7+ native security tools.

Repository: https://github.com/Armur-Ai/Pentest-Swarm-AI
Canonical: https://ross.abutalabs.com/products/pentest-swarm-ai
Language: Go
License: AGPL-3.0
License Family: copyleft
Topics: ai-agents, bug-bounty, cybersecurity, offensive-security, penetration-testing, penetration-testing-framework, penetration-testing-tools
Last push: 2026-08-26T15:48:40+00:00

## Health v2 (maintenance only)
Score: 75/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 99, release rhythm 51, longevity 63
- inputs: {"age_days": 890, "days_push": 7, "days_rel": 118, "gap_med": null, "n_releases_24m": 1}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2381, forks 446 (observed 2026-08-28T04:06:42.141732+00:00)

## What it is
An open-source autonomous penetration testing application that orchestrates a swarm of AI agents (recon, classification, exploitation, reporting) using ReAct reasoning and a shared stigmergic blackboard. Built in Go, it supports bug bounty, continuous monitoring, and CTF modes and works with Claude, OpenAI-compatible APIs, or fully local models via Ollama/LM Studio.

## Use cases
- automate penetration testing with ai agents
- run autonomous recon and exploitation on a target
- find and verify vulnerabilities for bug bounty hunting
- continuous security monitoring of my attack surface
- practice ctf challenges with an ai agent
- run pentesting fully locally with ollama
- generate pentest reports with evidence automatically

## When to choose
- you want autonomous, evidence-backed pentesting rather than passive scanning
- you need parallel coverage of large attack surfaces like thousands of subdomains
- you want to use local or self-chosen LLMs as the reasoning engine
- you run bug bounty, CTF, or continuous monitoring engagements

## When to avoid
- you need a mature, compliance-certified commercial pentest product
- you cannot legally authorize testing against your targets
- you require guaranteed accuracy — the project is alpha and LLM-driven findings need human review
- you need a lightweight passive scanner rather than an active exploitation harness

## Facets
- artifact type: application
- maturity: experimental
- function: agent-framework, penetration-testing, security, llm-inference, workflow-automation
- domain: security, penetration-testing, artificial-intelligence
- platform: go, cli
- tags: ai-swarm, offensive-security, bug-bounty, ctf, react-agents, autonomous-pentesting, stigmergic-blackboard, ollama, claude-api, ai-agents, linux, macos, docker

## Member repositories
- Armur-Ai/Pentest-Swarm-AI (main) score 75

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:42.141732+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T02:34:43.866092+00:00, confidence not recorded.
  - readme: https://github.com/Armur-Ai/Pentest-Swarm-AI (fetched 2026-08-28T04:06:42.141732+00:00, sha dbe4b30acd8c)
- Data as of 2026-08-30T08:39:29.467469+00:00.