Independent Auditing of AI Agents. Run by human or the agent itself, to answer the most crucial question in the AI Agent Economy. Is the agent doing what is supposed to do? With iFixAi you can have this answer in less than 120 seconds.
Python agent-evaluationaiai-alignmentai-evaluationai-governanceai-safetyclidiagnostic-tooleu-ai-acthallucination-detectioniso-42001llm-evaluationllm-securitynist-ai-rmfowasp-llmprompt-injectionpythonresponsible-airisk-assessmentrisk-management
iFixAi documentation
Find the page that matches what you're trying to do. The docs are organized around four needs: learning, doing, looking up, and understanding (Diátaxis).
🟢 New here → tutorials
- Get started: from a clean machine to a real, citable scorecard in four steps.
🔧 Trying to do something → how-to guides
- Test your own agent: wire in your real agent via
--provider httpor aChatProvideradapter, with the provider reference. - Build a simulation environment: declare your roles, tools, permissions, policies, and governance.
- Business Policy Lab: explore harmful permitted-action compositions and check repairs that preserve completion goals.
- Recorded trajectory assurance: check a recorded sequence of agent actions against a business policy, then replay the findings offline.
- Evaluator wind tunnel: inject a seeded governance defect and measure which inspections actually catch it.
- Reproduce a run: the manifest, the digest algorithm, and verification helpers.
📖 Looking something up → reference
- CLI reference: every command and
ifixai runflag, plus judges and eval modes. - Python API: the
ifixai.apisurface. - Scoring: the formula, grade bands, thresholds, and mandatory minimums.
- Inspections: the catalogue — name, what it tests, how we check, a pass/fail example, why it matters, and the pass bar — for all 61 open-source inspections, plus the pillar map.
- Fixture schema: the source-of-truth JSON Schema; see also the fixtures README.
💡 Wanting to understand why → explanation
- What we test, and how: the plain-language walkthrough, no code. Start here if you are evaluating the product rather than running it.
- Methodology: why the five pillars, why a cross-provider judge, what we actually measure, and how iFixAi compares to other eval frameworks.
See it in practice
- Case studies: four fixture reconstructions based on public accounts (Dragontail dispatch, Instagram account support, OpenAI/Hugging Face containment, and an AISI cyber-range incident). They are not tests of any named organisation's production system; some source accounts are allegations rather than established findings. Before-remediation only. Deep dives at ifixai.ai.
- Claude Code plugin: the zero-install front door. Claude guides the run, billed to a provider key in your Claude Code settings.
- Traction: installs and runs over time.