Parse → Validate → Execute

Your AGENTS.md is the spec.
We execute it.

The CI/CD platform for AI agents. Get your agent-readiness score in under 2 minutes—one file, no config.

Why execute vs. read-only?

Quick install

Get started in 60 seconds

One command, no install

npx @agentmd-dev/cli init

GitHub Action

Add to CI (Pinned SHA)

uses: iedayan/agentmd/.github/actions/agentmd@7f23caa9a688230815368fc07716828884479cad

Add to GitHub

Install the app

/github/install

npm package

Add to project

pnpm add @agentmd-dev/core

No install required · Free for 3 repos · No credit card

AGENTS.md
---
# Agent instructions for AI coding tools
---
 
## Build
 
```bash
$ pnpm run build
```
 
## Test
 
```bash
$ pnpm run test
```
 
## Lint
 
```bash
$ pnpm run lint
```
valid · score 94Markdown
Open source on GitHubBuilt for teams who ship with AI

60K+ repos

AGENTS.md standard

Governance-first

Guardrails & audit

2 min setup

No config required

Get your agent-readiness score

Paste your AGENTS.md or README below. No signup required—instant results.

185 / 50,000 chars
Results

Ready to validate

Click "Parse & validate" to see sections, commands, and your score.

Paste your AGENTS.md or press Ctrl+Enter to run

Why AgentMD?

The only platform that executes the AGENTS.md spec. Parse, validate, and run—with governance from the start.

One file, one spec

AGENTS.md is the emerging standard backed by Linux Foundation, OpenAI, and 60K+ repos. We execute it—no lock-in.

Governance built in

Guardrails, permissions, and audit trails from day one. Catch breaking changes before production.

Runs in your CI

GitHub Actions, push, PR, or cron. Same commands your agents use. Zero drift.

Why execute vs. read-only? →

Parse · Validate · Execute

A complete lifecycle engine for your AI agents, seamlessly integrated into your repository via Agentic UX.

Execute in CI

Run on push, PR, or cron. Results stream back via Chain of Thought logs. Catch breaking changes before they hit production.

agentmd run validate
Parsed AGENTS.md successfully
Executing reasoning chain... [10s]

Readiness score

Get a 0–100 score on how agent-ready your repository is.

98/100

Parse & validate

Extract sections and commands directly from your AGENTS.md file. Surface errors and syntax issues immediately in our dashboard before CI runs.

What you get

Agent-ready in three steps. Under a minute.

01

Create AGENTS.md

Run npx @agentmd-dev/cli init — auto-detects Node, Python, Rust, or Go. No install.

02

Add to CI

Drop the GitHub Action into your workflow. No install, no config.

03

Ship with confidence

Every run validated. Every command tracked. No drift.

Add to .github/workflows/ci.yml

.github/workflows/ci.yml
- uses: iedayan/agentmd/.github/actions/agentmd@7f23caa9a688230815368fc07716828884479cad
  with:
    command: validate
    path: .

Ready to ship with AI?

Get your score, add AGENTS.md, connect the GitHub App. Free for 3 repos—no credit card.

Ship with confidence. No drift, no surprises.