A repeated failure can become a regression fixture before it becomes a broad rule. Start with the smallest reproduction and a nearby example that should remain valid.

Failure → Minimal fixture → Negative case → Review rule
  1. 1Failure
  2. 2Minimal fixture
  3. 3Negative case
  4. 4Review rule

Work through the example

Prompt: Reduce this finding to synthetic source and identify a similar pattern that must not trigger.

One practical example

Use the prompt above on one small, named part of your app. Before accepting an edit, ask the agent to point to the source or configuration that supports its answer. If it cannot locate that context, resolve the missing input before expanding the task.

In the diagram, each arrow represents a handoff you can inspect. Keep the intermediate result rather than jumping directly to the final claim. For example, a reported finding should retain its location, and an executed check should retain its outcome. This gives the next attempt a concrete starting point and makes a misleading completion message easier to challenge.

Apply it to your project

Prompt: Reduce this finding to synthetic source and identify a similar pattern that must not trigger. Then compare the resulting diff with the original request. If it changed a neighboring concern, ask why that change was necessary. Save the result only after the relevant check has run, and record any part that remains unverified.

Evidence and limits

Do not upload private app code to build a public fixture.

This is an educational guide. Its presence in the series does not certify a completed client-specific lab. The series evidence record separates executed checks from exercises and blocked environments.

Inspect the source used in this lesson.

What to do next

Next: Verify Muse observer behavior before recommending it