// Guides

Guides

Problems anyone running AI coding agents against a real repository runs into, written from what this project’s own team ran into first. Each guide starts with the general approach, then shows how FULCRUMAXE handles it, and links to the pages and logs every claim comes from.

The guides

// Review

Stop an agent merging its own broken pull request

Merge gates an agent cannot set itself, what they catch, and where they still fail.

Read the guide →
// Safety

Run coding agents against a real repository safely

Credentials, isolation, gates and autonomy limits, with the limits stated plainly.

Read the guide →
// Cost

What a merged pull request actually costs

Median and 90th-percentile spend per merge from a real run, and what it leaves out.

Read the guide →
// Evidence

Catch an agent reporting work it never did

Check the record of what an agent did, not its report — and the detector built after two fabrications.

Read the guide →