How to Write, Test, and Publish a Reusable SKILL.md Agent Playbook (With Templates)
Step-by-step developer guide to authoring deterministic SKILL.md files, validating trigger descriptions, and listing on LightState.ai.
Learn the exact anatomy of a high-signal SKILL.md playbook—from YAML frontmatter trigger phrasing to verification scripts and anti-hallucination gates.
Verified Capabilities Benchmarked in This Guide
Inspect Live JSON Feed ↗Portable SKILL.md Authoring, Trigger Optimization & QA Playbook
Author, test, and package reusable `SKILL.md` playbooks compatible with Grok Bot, Claude Code, and Cursor: YAML frontmatter trigger phrasing, progressive d
Agentic Tool-Use Evaluation Harness & Prompt Grading Playbook
Design reproducible evaluation suites for tool-calling AI agents: golden test cases, deterministic rubric graders, latency/token cost tracking, and regress
Strict TypeScript AST & Runtime Zod Boundary Validation Skill
Enforce end-to-end type safety across API boundaries, environment variables, and LLM structured outputs using strict `tsconfig.json` flags and composable Z
Turborepo & pnpm Workspace Remote Caching & Task Graph Skill
Configure high-speed TypeScript monorepos with pnpm workspaces and Turborepo: deterministic task pipelines (`build`, `lint`, `test`), input hashing, and re
1. Anatomy of a Portable SKILL.md File
Every SKILL.md begins with YAML frontmatter containing a concise name and a trigger-rich description explaining WHEN the agent must load the skill, followed by numbered execution steps and verification commands.
2. Publishing & Claiming Your Skill on LightState.ai
Submit your GitHub repository or X launch post at /submit, then verify ownership via X OAuth 2.0 PKCE at /claim to unlock your Certified X Builder storefront (/u/<handle>).
Frequently Asked Questions
How long should a SKILL.md file be?
Keep the main SKILL.md under 500 lines for fast context loading, and place large reference tables in helper files.