AI readiness

See whether your repositories give AI agents the context and delivery pipeline they need to ship reliable work

Available at AI tools → AI readiness

AI coding agents are only as effective as the system they work in. They need two things from a repository:

• Instructions: configuration files such as AGENTS.md or CLAUDE.md that tell agents how to build, test, and navigate the codebase, and what "done" means.

• A fast path to production: agents open more pull requests than people do, so every change leans harder on your pipeline. When merged work waits hours to reach production, more AI-generated code only makes the queue longer.

AI readiness shows both for each active repository, so you can find the busy repositories where agents lack context or where changes are slow to ship.

What you see

A repository is active when it has merged pull requests in the selected timeframe. The summary at the top shows how many repositories are active, how many have a configuration file for at least one AI tool, and how many have full coverage.

For each active repository, the table shows:

• Time to deploy: the time from merge to production deployment for pull requests deployed in the selected timeframe. Use the aggregate selector to switch between average, median, and percentiles, and the comparison selector to compare against the previous period or Swarmia benchmarks.

• Merged PRs: pull requests merged in the selected timeframe. Use this to find the repositories where most work happens.

• Coverage: how many of the selected AI tools have a configuration file in the repository. Hover over the badge to see the breakdown per tool.

• One column per AI tool: hover over a configured tool to see the latest pull requests that updated its configuration files, or over an unconfigured one to see which files Swarmia looked for.

Filter the view by team, repository, and AI tool. When you select a team, merged PRs and time to deploy only include that team's pull requests.

How to use it

1. Start from the busiest repositories. Merged PRs tells you where agent context pays off the most. A frequently changed repository without configuration files is the first gap to close.

2. Check that instructions stay current. Hover over a configured tool to see when its files were last updated. Instructions written once and never touched tend to drift from how the codebase works.

3. Look at how fast changes reach production. A long time to deploy in an active repository means AI-generated work piles up after merge. Use CI visibility to find slow or failing workflows, and DORA metrics to follow delivery over time.

Where does the data come from?

The view is based on merged pull requests and the file paths reported by GitHub or GitLab. Configuration files are detected from pull requests, over all history and not just the selected timeframe. A file committed directly to a branch without a pull request, or merged before Swarmia started syncing your data, isn't detected. A file stays detected even after it's deleted.

Swarmia doesn't read your source code. Read more about data access.

The table only shows the AI tools enabled in your AI tool settings.

Time to deploy uses your production deployments. A repository without production deployments in the timeframe shows no value. A pull request deployed to production several times counts once per deployment.

Frequently asked questions

Why is time to deploy missing for a repository?

The repository has no production deployments with linked pull requests in the selected timeframe. Check that deployments are configured for it. See why time to deploy is 0 or missing for other common reasons.

Why is a repository missing from the table?

Only repositories with merged pull requests in the selected timeframe appear. If you've selected a team, the pull requests must belong to that team.