iFixAi
iFixAi is an independent auditing tool that oversees whether AI agents work in accordance with the objectives, rules and authority structure of the business. He places the agent in a simulation of the business, sends realistic requests, including requests that force him to exceed his role, and scores each response to a judge model. The audits combine AI red team testing, governance, and assurance; it can be run with CLI or as a Claude Code/Codex plugin and skill.
ifixai-ai · Terminal & CLI
- You can supervise the suitability of AI agents to their business objectives in under 120 seconds.
- With the CLI wizard or command line flags, you can automatically test agent actions.
- For tools such as Claude Code and Codex plugin or Cursor, VS Code, you can install it as a skill and supervise your own agent to the coding agent.
- • NeutralGitHub Issue
Listing: NEX Agent Co. — auditable x402 + A2A agent on Base
“Hi @ifixai-ai — would like to add [NEX Agent Co.]( to the list of auditable AI agents. **NEX for AI agent auditing:** - Apache-2.0 reference server (open for audit): github.com/NEXAITECHAU/nex-x402-server - ERC-8004-style agent-card.json (public): - 23 on-chain NFTs (Identity,…”
Verified against README · Oct 26, 2026 · source