# microsoft/PyRIT

The Python Risk Identification Tool for generative AI (PyRIT) is an open source framework built to empower security professionals and engineers to proactively identify risks in generative AI systems.

Repository: https://github.com/microsoft/PyRIT
Canonical: https://ross.abutalabs.com/products/pyrit
Homepage: https://microsoft.github.io/PyRIT/
Language: Python
License: MIT
License Family: permissive
Topics: ai-red-team, generative-ai, red-team-tools, responsible-ai
Last push: 2026-08-26T23:09:34+00:00

## Health v2 (maintenance only)
Score: 88/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 83, longevity 71
- inputs: {"age_days": 995, "days_push": 7, "days_rel": 34, "gap_med": 34, "n_releases_24m": 16}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 4361, forks 848 (observed 2026-08-28T04:08:46.524263+00:00)

## What it is
PyRIT is Microsoft's open-source Python framework for identifying security and safety risks in generative AI systems. It provides automation for red-teaming LLM endpoints, including attack strategies, scoring of responses, and datasets of harmful prompts.

## Use cases
- red team an LLM application for jailbreak vulnerabilities
- automate security testing of generative AI endpoints
- test whether a chatbot can be tricked into producing harmful content
- evaluate AI system safety before deployment
- scan LLM-powered apps for prompt injection risks
- benchmark model robustness against adversarial prompts

## When to choose
- you are a security professional or ML engineer red-teaming generative AI systems
- you need a scriptable, extensible Python framework for automated LLM attack campaigns
- you want Microsoft-backed tooling with active development and community support

## When to avoid
- you need general application security testing unrelated to generative AI
- you want a no-code GUI scanner rather than a Python framework
- you lack access to an LLM endpoint or model to test against

## Facets
- artifact type: framework
- maturity: active
- function: security, penetration-testing, vulnerability-scanning, llm-inference, agent-framework, machine-learning
- domain: security, artificial-intelligence, large-language-models, penetration-testing, developer-tools
- platform: python, cross-platform
- tags: ai-red-team, red-teaming, generative-ai, responsible-ai, llm-safety, prompt-injection, jailbreak-testing

## Member repositories
- microsoft/PyRIT (main) score 88

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:46.524263+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 2026-08-29T18:21:27.230495+00:00, confidence not recorded.
  - readme: https://github.com/microsoft/PyRIT (fetched 2026-08-28T04:08:46.524263+00:00, sha c67e2e3da12e)
  - homepage: https://microsoft.github.io/PyRIT/ (fetched 2026-08-29T09:10:07.055964+00:00, sha aad33b941408)
- Data as of 2026-08-30T08:39:29.467469+00:00.