# rogue-security/rogue

AI Agent Evaluator & Red Team Platform

Repository: https://github.com/rogue-security/rogue
Canonical: https://ross.abutalabs.com/products/rogue
Homepage: https://qualifire.ai
Language: Python
License: NOASSERTION
License Family: other
Topics: agents, ai, ai-agents, e2e-testing, llm, testing, testing-framework
Last push: 2026-08-04T07:52:34+00:00

## Health v2 (maintenance only)
Score: 78/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 96, release rhythm 81, longevity 32
- inputs: {"age_days": 454, "days_push": 29, "days_rel": 126, "gap_med": 2, "n_releases_24m": 34}
- flags: no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1058, forks 162 (observed 2026-08-28T04:03:25.065585+00:00)

## What it is
Rogue is an open-source AI agent evaluator and red teaming platform that stress-tests agents against business policies and simulated adversarial attacks. It supports A2A, MCP, and direct Python function agents, with a server, terminal UI, and CLI for CI/CD integration.

## Use cases
- red team my AI agent for security vulnerabilities
- test llm agent against business policies
- run adversarial attack simulations on chatbot
- check agent compliance with OWASP LLM top 10
- regression test ai agent behavior in CI pipeline
- evaluate MCP server agent for jailbreaks
- generate compliance reports for EU AI Act

## When to choose
- you need automated adversarial testing of AI agents before production
- you want policy compliance and behavior regression testing with pass/fail reports
- your agent speaks A2A, MCP, or is a plain Python function
- you need CVSS-scored vulnerability findings mapped to compliance frameworks

## When to avoid
- you need runtime guardrails or real-time request blocking rather than pre-production testing
- you want general-purpose unit testing of non-agent Python code
- you need LLM observability and tracing in production, which the parent Qualifire platform handles

## Facets
- artifact type: framework
- maturity: active
- function: testing, e2e-testing, agent-framework, security, penetration-testing, llm-inference, cli
- domain: security, testing, large-language-models, developer-tools
- platform: python, cli, cross-platform
- tags: red-teaming, llm-evaluation, ai-agent-testing, adversarial-attacks, owasp-compliance, mcp, a2a-protocol, tui, ci-cd, ai-agents, docker

## Member repositories
- rogue-security/rogue (main) score 78

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:25.065585+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 2026-08-30T06:57:52.216731+00:00, confidence not recorded.
  - readme: https://github.com/rogue-security/rogue (fetched 2026-08-28T04:03:25.065585+00:00, sha 7520c90a8dd7)
  - homepage: https://qualifire.ai (fetched 2026-08-29T12:59:47.682668+00:00, sha b0bb8e9e9a08)
  - site_page: https://docs.qualifire.ai/ (fetched 2026-08-29T12:59:47.687389+00:00, sha 2df28aab2811)
  - site_page: https://qualifire.ai/about (fetched 2026-08-29T12:59:47.690565+00:00, sha fc7b37c56b77)
  - site_page: https://docs.qualifire.ai/introduction (fetched 2026-08-29T12:59:47.692086+00:00, sha 7cc4e8a1eff4)
  - site_page: https://qualifire.ai/product (fetched 2026-08-29T12:59:47.685523+00:00, sha 2b73336c0d46)
  - site_page: https://qualifire.ai/pricing (fetched 2026-08-29T12:59:47.689043+00:00, sha 0cc9b91a1681)
- Data as of 2026-08-30T08:39:29.467469+00:00.