repowiseifixai-ai/ifixai

ifixai-ai / ifixai

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★ 33,897
Use in your agentAsk this codebase

github.com/ifixai-ai/ifixaiIndexed at 92bb5d0Up to date with upstream

What is ifixai?

ifixai is a Python tool for independently auditing AI agents for operational alignment and assurance. It simulates business requests, including requests that test an agent's role boundaries, and has a judge grade each response. The project is implemented primarily in Python and uses a command-line interface with provider integrations.

Written by a model from the README and the source

864 files and 7,710 symbols in 5 modules, led by Python (610 files) and Shell (1).

Start reading

Code health

8.4out of 10Good

Hotspots average 6.6, the place fixes pay off most.

Full health report

On the leaderboards:#3 of 50 Python repos

Use ifixai-ai/ifixai in your agent

Point Claude Code, Cursor, Codex or VS Code at this repository. Your agent gets the index this page is built from: cited answers, callers, history and health for any file.

  1. 1.In Claude, open Customize → Connectors, click +, then Add custom connector. On a Team or Enterprise plan an owner adds it first under Organization settings → Connectors.
  2. 2.Paste the server URL below as the Remote MCP server URL, then click Add.
  3. 3.Click Connect. A Repowise sign-in window opens. Approve access.
  4. 4.Repowise's tools are now available in your Claude chats.

Connect with one click. Sign in with Repowise (OAuth), no API key to copy. Works in Claude on the web, desktop, and mobile.

Docs
bash
https://api.repowise.dev/mcp/ifixai-ai/ifixai
Editors need an API key: sign in to create one.Sign in to connect

This repo is public, so anyone with a Repowise account and an API key can connect.

Try asking

  • get_overview()
  • get_answer("how is ifixai put together?")
  • get_why("why is ifixai structured this way?")

Explore ifixai

Key features of ifixai

  • ifixai simulates business requests that test an AI agent's intentions, goals, permissions, and role boundaries.
  • ifixai grades agent replies and provides evidence showing what was requested and what the agent did.
  • ifixai combines AI red teaming, governance, assurance, and philosophy in its inspections.
  • ifixai produces pass or fail results, an A to F grade across five core pillars, and JSON and Markdown reports.
  • ifixai can audit OpenAI-compatible chat endpoints and supports judges from providers including OpenAI, Anthropic, Gemini, Azure, Bedrock, and OpenRouter.

How ifixai-ai/ifixai works

An audit starts through ifixai/cli/main.py or ifixai/cli/run.py, which provide the command-line entry points for authoring profiles, validating fixtures, and running inspections. The agent profile defines the simulation environment and the requests sent to the agent. Provider integrations in ifixai/providers/base.py connect the audit to the agent and to a judge selected from a provider. ifixai/evaluation/analytic_judge.py evaluates the agent's replies against the inspections. ifixai/core/types.py defines shared data structures used across the audit flow. ifixai/reporting/scorecard.py produces the scorecard and supports JSON and Markdown reports.

The codebase has 864 files and 7,710 symbols in 5 modules, led by Python and Shell. Code health is 8.4 out of 10, rated good.

Getting started with ifixai

  1. Create a virtual environment and install ifixai.
    pip install ifixai
  2. Create an example agent reference profile.
    ifixai author reference -o reference.yaml
  3. Validate the profile.
    ifixai validate reference.yaml
  4. Run the demo audit.
    ifixai run --provider mock --eval-mode self --fixture reference.yaml

Commands copied from the README

ifixai-ai/ifixai FAQ

What is ifixai?

ifixai is an independent auditor for AI agents. It tests whether an agent follows its intended role, goals, permissions, and policies.

How does ifixai work?

ifixai places an agent in a simulated business environment, sends it operational requests, and has a judge grade each reply. It produces findings, scorecards, and reports.

How do I install ifixai?

Install ifixai with pip install ifixai, then create and validate a reference profile before running an audit. Python 3.10 or newer is required.

What language is ifixai written in?

ifixai is written primarily in Python. Its package includes command-line entry points and provider integrations.

How is the ifixai code organized?

ifixai is organized mainly under the ifixai directory, with CLI code, evaluation, reporting, core types, and provider modules. The main entry points include ifixai/cli/main.py, ifixai/cli/run.py, and plugin/hooks/bootstrap.py.

Who is ifixai for?

ifixai is for people who need to test an AI agent's operational alignment and assurance. It supports audits of agents with chat endpoints, including OpenAI-compatible endpoints.

What license does ifixai use?

ifixai uses the Apache-2.0 license.

The summary, features and answers on this page are written by a model from the README and the source, and refreshed with each index.

About this page

Get a page like this for your own codebase

Repowise built this map of ifixai-ai/ifixai from its source and git history. It parses every symbol, draws the dependency graph, scores each file's health from complexity, duplication, coverage and churn, finds hotspots and owners in the history, and writes documentation for every module. Your coding agent reads the same index over MCP, so its answers cite real files.

Index your repo freeSee how Repowise worksFree for 2 public repos, with a page like this one.

Prefer local? pip install repowise && repowise init indexes your own checkout, no account needed.