How can we help?

aictrl uses generative AI to help engineering teams review code, understand pull requests, and run configured development workflows. Treat generated findings, explanations, suggestions, and code as material for a human to review.

AI can miss defects, invent details, or misread incomplete context. A source citation or a grounded-claim count is a way to inspect the evidence, not proof that a conclusion is correct. Check the referenced commit and code, run relevant tests, and review security and privacy implications before acting on an output or merging a change. Workflow results depend on the configured skills, tools, model, permissions, and available context.

Use these features to support engineering decisions. Do not rely on AI output alone for security assurance, legal or compliance decisions, or decisions about hiring and employee performance. Review workflow permissions and approval gates before starting a run.

To report an incorrect, unsafe, or unwanted output, contact us with the affected feature, approximate time, expected result, and a redacted description. Include a run or pull-request link only if you are allowed to share it. Do not send credentials, private source, personal data, or confidential organization information. You can report a problem by email without opting into product analytics or using an in-app survey. For a concern about an app distributed through GitHub Marketplace, you can also contact GitHub Support.

If you believe you have found a vulnerability, email a concise reproduction and impact assessment. Do not access other customers’ data, degrade service, or publish the issue before we have had a reasonable opportunity to investigate.