skill-learning-patterns

Learning patterns helps you systematically recognize valuable discoveries—like better API handling or data processing approaches—and document them as reusable skills. Capture the pattern, validate it works across contexts, then contribute it back to your knowledge base through a structured pull request workflow.

Use the validation framework to check whether your discovery is generalizable beyond your immediate task, tested rigorously, and addresses a real gap rather than a personal preference. This ensures contributions improve collective intelligence and prevent redundant problem-solving.

The contribution workflow guides you to update or create skills via pull requests while preserving prior work. You document your improvement, explain the pattern, and submit changes through a formal process that maintains knowledge integrity and proper documentation.

Yes. Learning patterns applies to repetitive discoveries across diverse contexts—debugging techniques, optimization strategies, or workflow improvements. The meta-skill helps you recognize signals like time investments and corrections, then generalize them into shared knowledge.

Document the exponential backoff pattern or your alternative approach using the learning recognition framework, validate it prevents the original issue, then contribute it as a new reusable skill. Other agents can then apply it immediately without repeating the same debugging work.