Ship AI agents with confidence
Analyze agent source code
AI-powered analysis finds misconfigurations, risky tool access, missing guardrails, and other issues. Automatically triage false positives.
Evaluate agents at runtime
35+ built-in evaluations. Check for factual accuracy, instruction adherence, prompt injection, jailbreaks, and more. Framework-agnostic.
The AgentOps platform
The CLI finds problems. The platform fixes them. Get auto-fix PRs and CI/CD integration.
FAQs
What does Flint AI do?
Flint AI is a free CLI that tells you whether your AI agent is reliable before you ship it.
flintai scan runs analysis on your agent's Python source code. It checks for security vulnerabilities, misconfigurations, and quality issues. flintai eval sends adversarial and functional prompts to your running agent and scores the responses.
Get a reliability score per agent, with findings mapped to OWASP ASI01–ASI10 and scored with CVSS v4.
Two commands, one CLI. Works with any major agent framework.
What do I need to run the Flint AI CLI?
Python 3.13+ and an LLM API key. The key powers the AI reasoning in flintai scan and the prompt generation and scoring in flintai eval. OpenAI, Anthropic, and Google Gemini keys are supported. Gemini 2.5 Flash has a free tier, so you can start without spending anything. Install with pip install flintai-cli, add your key to a .env file, and run your first scan.
Does code leave my machine?
No. Flint AI runs locally. The only data that leaves your machine is what gets sent to the LLM provider whose API key you supply. Flint AI CLI has no backend, no telemetry server, and no account system. Your source code stays on your machine.
Does it cost anything?
No. The CLI is free with no usage limits or trial period. You bring your own LLM API key.