reflect-appworld-failure

Analyze AppWorld task failures by extracting actionable patterns from error messages, failed code snippets, and root causes. This Skill generates structured JSON bullets with specific code examples, tags, and confidence scores that capture the core mistake and its fix, turning each failure into a reusable debugging pattern.

Yes. When AppWorld tasks fail with API errors, this Skill analyzes the error type, error messages, and failed code to identify root causes and generate targeted playbook bullets. Each entry includes working code examples and evidence so you prevent the same mistake in future tasks.

Provide the task description, affected apps, error type classification, error messages, the failed code snippet, missing code patterns, and suggested fixes. The Skill processes these sections to produce structured learning bullets with fields for id, title, content, tags, evidence, confidence, and scope.

Pattern extraction identifies root causes and common coding mistakes from each failure, then documents them with concrete code examples and tags. By building a permanent playbook of patterns tied to specific error types and apps, you avoid re-solving the same problem across different tasks.

Error analysis identifies why a specific task failed by examining error messages and code; pattern extraction generalizes that failure into a reusable coding pattern with working examples. Together, they transform one failure into a learning asset applicable across multiple AppWorld tasks.