Radar / AI security / iFixAi
iFixAi: independent auditing of AI agents
Independent AI-agent auditor: 250+ inspections across 69 misalignment categories (red teaming + operational assurance) returning an A-F grade with evidence in under 120 seconds.
Why it matters
First open, judge-independent audit harness for agents: CLI wizard, Claude Code/Codex plugin or agent-run skill, cross-vendor judge pairing (never self-graded), and compliance mapping to EU AI Act, NIST AI RMF, OWASP LLM and ISO 42001.
What you could build with it
A company deploying customer-facing agents could run iFixAi audits continuously against production traces to catch drift from intended behavior. The result: an early-warning system for agent misbehavior that reports in plain language a manager can act on.
Does it hold up?
Usable and independently exercised: pip-installable with a free mock run, 26,641 stars, and one independent developer documented fixing a real audit finding iFixAi surfaced; the 120-second budget keeps coverage shallow, so pair with runtime observability for high-stakes agents.
Built with iFixAi
- openwatari: audit trail fix driven by an independent iFixAi rungithub.com · Independent developer ran iFixAi against his own agent (openwatari): the audit scored audit-record completeness at 0%, and he rewired every audit line to carry actor, decision and rule_applied - a real finding driving a real fix.
- snehu34/ifixai - community plugin skill usage notesgithub.com · Third-party repo wrapping the iFixAi plugin SKILL.md with usage guidance (dry-run cost estimates, single vs multi-judge setups) - community operational documentation.
- iFixAi on Product Huntproducthunt.com · Product Hunt launch page (12h old) with founder Q&A on simulation environments, authority boundaries and prompt-injection coverage, plus a 1,000-free-audit promo code PH1000IF.
- ifixai-ai/ifixai-plugingithub · Official Claude Code marketplace plugin: sign in from inside Claude Code, connect your deployed agent, get red-teaming plus operational-assurance findings back as a report.
- ifixai-ai/ifixai-connectorgithub · Companion connector repo for the iFixAi Claude Code plugin (MCP server pointer, skill, manifest).
Learn more
First spotted on github: source.
More AI security
OpenAI, Anthropic and Google DeepMind jointly unveil cyber-focused safety models and safeguardsThe three rival labs disclosed weeks of behind-the-scenes coordination and jointly released a set of cyber-focused AI…security · JEV 0.73OpenAI textGrain watermarking for ChatGPT and Codex in the EUOpenAI will roll out its invisible textGrain watermarking system to ChatGPT/Codex users in the EU over the coming…security · JEV 0.65Cloudflare security-audit-skill: multi-phase security audits for coding agentsA coding-agent skill that runs multi-phase security audits with independently verified, machine-checkable findings.security · JEV 0.63LiveNerf: a pre-registered benchmark for post-release model driftOpen-source, pre-registered 30-day benchmark that detects whether Claude Opus 5.5 quietly gets worse after launch.security · JEV 0.61Sandlock 0.8.8: deferred commit for agent sandboxesThe process-based Linux AI sandbox (no container, no VM) ships deferred commit: every run returns a changeset of what…security · JEV 0.58Microsoft Execution Containers (MXC) goes GA on Windows 11At its Oct 7 Windows AI event with NVIDIA, Microsoft announced general availability of Microsoft Execution Containers…security · JEV 0.58
Get the week's best AI launches, plus 3 ideas worth building
One email every Saturday. Ranked by traction, not hype. Free.