How well can coding agents work in your repo?

Paste a public GitHub repository. Get a readiness score and your biggest problem, with evidence from the actual repo — free.

Private repo?

Paste a fine-grained GitHub token with read-only access to just this repository. It's used for this one analysis and never stored.

Free score · no signup · works on private repos · full remediation report $49

The full report · $49

Every fix, ranked — plus your AGENTS.md, generated.

The free score shows you the biggest problem. The full report is the remediation plan: everything a coding agent needs your repo to say, written for your repo specifically.

Five-dimension breakdown

Environment & setup, testing & verification, architecture context, agent instructions, and tool discoverability — each scored /20 with the reasoning.

Evidence-based findings

Every finding cites the actual files, scripts and commands in your repo — plus its impact on a coding agent and the exact change to make. No generic advice.

Generated AGENTS.md

A repository-specific AGENTS.md (and optional CLAUDE.md) built from what's really in your repo — drop it in the root and agents read it automatically.

How it works

No clone, no OAuth, nothing stored.

Paste your repo. Agent Ready reads the tree, README, manifests, CI config and agent instruction files through the GitHub API.
Claude evaluates it. One question drives the whole rubric: can a coding agent work here without constantly guessing?
Get the score free. Readiness out of 100, your verdict, and the single biggest problem with evidence.
Unlock the fix plan. $49 for every finding, the prioritized fixes, and your generated AGENTS.md — as a page and a PDF.

Answer engine

Agent Ready, answered directly.

Short answers for teams making their repositories work well with coding agents.

What is an agent-ready repository?

An agent-ready repository is one a coding agent (Claude Code, Codex, Cursor, GitHub Copilot) can work in without guessing: setup and run commands are documented, there is a clear validation sequence that defines when a change is done, major directories are explained, and agent instructions (AGENTS.md or CLAUDE.md) capture repository-specific rules and gotchas.

How does the Agent Ready score work?

Agent Ready reads your repository through the GitHub API — the tree, README, manifests, CI config and agent instruction files — and scores five dimensions out of 20: environment and setup, testing and verification, architecture context, agent instructions, and tool discoverability. Every finding must cite evidence from the repository itself; the free score includes your biggest problem in full.

What do I get in the full report?

The $49 full report includes the complete five-dimension breakdown with reasons, every finding with evidence, impact and the exact fix, a prioritized fix-first plan, a repository-specific AGENTS.md generated for your repo, an optional CLAUDE.md for Claude Code, and a print-ready PDF.

Does Agent Ready need access to my code?

Agent Ready reads public repositories through the GitHub API — no cloning, no OAuth, no accounts, and nothing is stored. Private repository support is coming; for now, run it against a public mirror or contact us.