Your assistant writes the AL. Who checks it?
Claude Code is a great way to write Business Central code. It is not a team review gate, and it cannot be an independent check on code it wrote itself.
Where Claude Code is enough
If you are one to ten developers and one person can still hold the codebase and the rule set in their head, an assistant plus your own conventions works. Honestly, you may not need PEQY yet.
That stops being true the moment review has to work as a system instead of a person. Three things break first:
- An assistant per developer is not a review gate. Everyone prompts differently; nothing guarantees every pull request gets the same scrutiny before it lands.
- The author cannot be the check. When the same model writes and reviews the code, its blind spots pass straight through. Review only works when it is independent of the author.
- Nobody owns the rules. Personal prompts drift, differ per developer, and leave with the person who wrote them.
What an independent, AL-native reviewer adds
- A second pair of eyes that did not write the code. PEQY reviews every pull request the same way, whoever (or whatever) wrote it.
- Findings verified before you see them. Every compile error PEQY claims is checked against the actual AL compiler first.
- Grounded in the Business Central symbol model. Base app, system app and platform symbols are resolved, not guessed from the diff.
Frequently asked questions
Is Claude Code enough for a small Business Central team?
Often, yes. With one to ten developers and one person who can hold the codebase and the rule set in their head, an assistant plus your own rules works. PEQY becomes worth it when review has to function as a team gate: when no single person sees every pull request anymore.
Can the assistant that wrote the code also review it?
Not independently. A model reviewing its own output shares its own blind spots: whatever it got wrong writing the code, it tends to accept when reading it. PEQY is a separate reviewer that verifies claims against the AL compiler and the Business Central symbol model instead of trusting the diff.
Does PEQY replace Claude Code or Copilot as an assistant?
No. Keep whatever your developers write code with. PEQY sits on the other side of the pull request as the review gate. It checks what lands, regardless of who or what wrote it.
Which platforms does PEQY support?
GitHub, GitLab (cloud and self-hosted) and Azure DevOps, including on-premises Azure DevOps Server and TFVC. Hosted in Germany, in the EU.
Keep your assistant. Add the gate.
Point PEQY at a repository and let it review your past pull requests. You judge the findings.