$ show alchaincyf/darwin-skill
Darwin-skill is a meta-skill (SKILL.md) and supporting tooling for autonomously optimizing agent skills (SKILL.md files). It implements a ratchet-based hill-climbing loop that evaluates skills across 9 dimensions (structure, effectiveness, meta-skill), spawns independent LLM judges for paired comparison, and only retains improvements validated by human-in-the-loop checkpoints. The repo includes a Node.js screenshot script for generating result cards, HTML templates, documentation, and references integrating Microsoft Research's SkillLens and SkillOpt papers. Despite being primarily a markdown document, it exhibits high craftsmanship: rigorous empirical grounding, explicit anti-pattern blacklist, versioned design philosophy, and detailed operational procedures. The actual executable code is minimal (one utility script), but the skill definition itself is a sophisticated artifact that demonstrates deep understanding of LLM evaluation pitfalls and optimization strategies.
The LLM identified 1 files worth highlighting for reuse:
A Node.js script that uses Playwright to capture a screenshot of a specific HTML element (.card) at 2x resolution. It includes robust fallback logic to locate playwright/playwright-core globally or via npm root, handles cross-platform file opening (Windows/macOS/Linux), and prints dimensions. This pattern (parameterized, fallback-aware, self-contained) can be extracted and reused in any project requiring automated HTML element screenshots with error-resilient dependency resolution.