$ show DietrichGebert/ponytail
Ponytail is a skill ecosystem for AI coding agents that enforces a 'lazy senior developer' mindset through a priority ladder: YAGNI, reuse existing code, prefer stdlib/native features, avoid unrequested abstractions, and write minimal working code. It delivers this via platform-specific adapters (hooks, command files, plugin manifests) for agents like Claude Code, Codex, GitHub Copilot, OpenCode, Gemini CLI, and others. The core skill definition lives in skills/ponytail/SKILL.md, with companion skills for review, audit, debt tracking, gain scoring, and help. The system includes automated validation scripts to maintain consistency across platform-specific rule copies and version manifests, benchmark suites to measure code/token/cost reductions, and test coverage for hook behavior and plugin integrity.
The LLM identified 3 files worth highlighting for reuse:
Verifies that platform-specific rule files (Cursor, Windsurf, Cline, etc.) match the canonical AGENTS.md body. Uses fs.readFileSync and string comparison to detect drift, exits with error code on mismatch. Self-contained, dependency-free, and applicable to any project needing to keep multiple rule copies in sync with a source of truth.
Ensures version consistency across package.json and all plugin manifests (Claude, Copilot, Gemini, OpenCode, etc.). Extracts version fields via JSON parsing and compares against package.json version. Demonstrates defensive programming with detailed error reporting and exit codes. Reusable for any multi-manifest project requiring version lockstep.
Provides writeHookOutput() and getHookSpecificOutputKey() to abstract platform-specific hook response formats (Claude raw text, Codex/Qoder hookSpecificOutput, Cursor JSON). Eliminates duplication in activation/mode-tracker/subagent hooks. Isolated pattern with minimal internal dependencies (only fs/path) that could be extracted as a universal hook adapter for any agent platform.