How it works

Foe: Flaky Wisp, a failing or flaky test.

Flaky Wisp

1. Flaky Wisp

   The Flaky Wisp blocks the way.

   What really happensIt stands for a failing or flaky test.

   Sentinel: Context 100/100.

2. Flicker

   The Flaky Wisp uses Flicker.

   What really happensA flaky test passes on one run and fails on the next.

   Sentinel takes 8. Sentinel: Context 92/100.

3. Unit Testing (wshobson agents)

   The hero raises the Unit Testing relic over untested code.

   What really happensYou run /unit-testing:test-generate and describe the code or module to cover.

   Flaky Wisp takes 29.

4. Unit Testing (wshobson agents)

   It counts the open flanks the wisp could slip through.

   What really happensThe command looks for untested functions and classes and picks pytest for Python or Jest for JavaScript and TypeScript.

   Flaky Wisp takes 24.

5. Flicker

   The Flaky Wisp uses Flicker.

   Sentinel takes 8. Sentinel: Context 84/100.

6. Unit Testing (wshobson agents)

   Wards rise around each path: mocks, edge cases, clear checks.

   What really happensIt writes tests with mocked dependencies, edge cases and assertions, then reports coverage gaps.

   Flaky Wisp takes 36.

7. Unit Testing (wshobson agents)

   When a test turns red, the debugger ally hunts the cause.

   What really happensThe unit-testing-debugger agent captures the error, reproduces it, isolates the failure, makes a minimal fix and checks it.

   Debugger: Context 40/40.

8. Unit Testing (wshobson agents)

   The summons work the problem, each in its own context.

   What really happensEach subagent does its own model and tool work, apart from the main conversation.

   Flaky Wisp takes 11. Flaky Wisp falls.

9. Unit Testing (wshobson agents)

   Debugger returns with a short report.

   What really happensOnly the subagent's final summary comes back to the main agent.

   Sentinel takes 2. Sentinel: Context 82/100. Sentinel: Insight. Debugger returns.

10. Victory

    The Flaky Wisp falls.