Project AGENTS Init
Initializes and audits concise AGENTS.md files and root documentation to optimize AI context usage and support cross-session recovery.
Project AGENTS Init establishes lightweight, standardized AGENTS.md guidelines and centralized docs/ hierarchies for software repositories. Designed to prevent context bloat and minimize AI token overhead, it emphasizes delegating enforcement to linters and CI rather than verbose prose. The skill also introduces standardized execution tracking files, allowing developers to pause work and seamlessly resume complex tasks across new sessions or different machines without losing critical architectural context.
Características Principales
01Initializes and audits concise, token-efficient AGENTS.md project guidelines
02Establishes a lightweight execution progress format for cross-device session recovery
03Evaluates command accuracy and token expenditure before applying documentation updates
04Prefers automated linting, typing, and CI checks over redundant written instructions
05Centralizes repository architectural notes and specifications within root docs/
060 GitHub stars
Casos de Uso
01Tracking and resuming multi-step project tasks across distinct coding sessions or devices
02Auditing and trimming bloated AGENTS.md files to reclaim prompt context window space
03Setting up structured, low-overhead guidelines for new repositories worked on with Claude Code