skills AGENTS.md

A documentation-based guide for creating reusable instructions for coding agents, including generated, synced, and hand-written skills.

In plain words
What is it for?
Use it to generate skills from projects such as Vue, Nuxt, Vite, and UnoCSS, or sync existing skills from projects such as Slidev and VueUse.
Why use it?
It helps turn project documentation into focused agent instructions without copying irrelevant user guides or creating unnecessary reference material.

Instructions file for CodexOpenCode

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add instructions/antfu/skills/agents-md
Clone the repo
git clone --depth 1 https://github.com/antfu/skills

Made for: Codex, OpenCode.

Per session 1,919 This file is loaded in full into every session.
When invoked 1,919 The same file — it is already loaded in full.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.01919 $0.01919
Opus 5 $0.00959 $0.00959
Sonnet 5 $0.00384 $0.00384
Haiku 4.5 $0.00192 $0.00192

Measured 3d ago against content hash 874b156c9a0f, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

skills AGENTS.md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 3d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

Origin

Copies of this mod

2 near-identical copies found in the catalogue:

AGENTS.md · 254 lines

How it starts

The opening of the file, as written. The whole thing — 254 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Skills Generator

Generate Agent Skills from project documentation.

PLEASE STRICTLY FOLLOW THE BEST PRACTICES FOR SKILL: https://platform.claude.com/docs/en/agents-and-tools/agent-skills/best-practices

  • Focus on agents capabilities and practical usage patterns.
  • Ignore user-facing guides, introductions, get-started, install guides, etc.
  • Ignore content that LLM agents already confident about in their training data.
  • Make the skill as concise as possible, avoid creating too many references.

Skill Source Types

There are two types of skill sources. The project lists are defined in meta.ts:

Type 1: Generated Skills (sources/)

For OSS projects without existing skills. We clone the repo as a submodule and generate skills from their documentation.

  • Projects: Vue, Nuxt, Vite, UnoCSS
  • Workflow: Read docs → Understand → Generate skills
  • Source: sources/{project}/docs/

Type 2: Synced Skills (vendor/)

For projects that already maintain their own skills. We clone their repo as a submodule and sync specified skills to ours.

  • Projects: Slidev, VueUse
  • Workflow: Pull updates → Copy specified skills (with optional renaming)
  • Source: vendor/{project}/skills/{skill-name}/
  • Config: Each vendor specifies which skills to sync and their output names in meta.ts

Type 3: Hand-written Skills

For skills that are written by Anthony Fu with his preferences, experience, tastes and best practices.

You don't need to do anything about them unless being asked.

Repository Structure

.
├── meta.ts                     # Project metadata (repos & URLs)
├── instructions/               # Instructions for generating skills
│   └── {project}.md            # Instructions for generating skills for {project}
│
├── sources/                    # Type 1: OSS repos (generate from docs)
│   └── {project}/
│       └── docs/               # Read documentation from here
│
├── vendor/                     # Type 2: Projects with existing skills (sync only)
│   └── {project}/
│       └── skills/
│           └── {skill-name}/   # Individual skills to sync
│
└── skills/                     # Output directory (generated or synced)
    └── {output-name}/
        ├── SKILL.md           # Index of all skills
        ├── GENERATION.md       # Tracking metadata (for generated skills)
        ├── SYNC.md             # Tracking metadata (for synced skills)
        └── references/
            └── *.md            # Individual skill files

Read the full file on GitHub · 254 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 3d ago First seen · 254 lines · 1,919 tokens per session scan A 874b156c9a0f

Subscribe to this mod's changes

skills AGENTS.md is an instructions file published in the GitHub repository antfu/skills (5,826 stars, last pushed 2mo ago), licensed MIT. It adds 1,919 tokens to every session, about $0.0096 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.