Implements reflexion loop for self-correcting deliverables: implement, validate (tests/security/accessibility), self-critique issues, retry up to 3 iterations. Ensures quality in code/content/AI/services.
This skill uses the workspace's default tool permissions.
Preview
Self-correcting implementation loop from the n-trax pattern.
SKILL.md
Similar Skills
doubt-driven-development
37.4k
Subjects non-trivial code decisions to fresh-context adversarial review before finalizing. Use for high-stakes scenarios like production changes, security logic, or unfamiliar codebases.
Orchestrates implement-analyze-fix loops: implements code, AI-reviews changes, fixes issues, repeats until clean or max iterations. For iterative development with quality checks.
The auto-dogfood system is an inferential verification loop
Order: Always attempt rules-based → computational → inferential. Only escalate to the next mode when the previous mode cannot verify the property in question.
Source: Trivedy (Anatomy of an Agent Harness, LangChain blog). Three-mode taxonomy adapted from Böckeler (Harness Engineering, martinfowler.com — computational vs inferential distinction). Note: harnesses continue to matter even as models improve — they engineer systems around model intelligence, not just patch deficiencies.
Rules
Each iteration must show measurable improvement over the previous.
If the same issue recurs across iterations, investigate root cause rather than patching symptoms.
Never skip the self-critique step, even if tests pass.
Log the reflexion loop outcome in delivery-journal.md.