skill-from-github

A workflow for creating a coding skill from well-maintained GitHub projects. It searches for projects that solve a requested task, checks their quality, and uses what it learns to shape the skill.

In plain words
What is it for?
Use it to find libraries or tools for tasks such as converting Markdown to PDF, analysing reviews, or generating API documentation, then turn that research into a reusable skill.
Why use it?
It helps when a task needs knowledge that is not already available in the agent’s skills. The quality checks reduce the chance of basing the skill on abandoned or incomplete code.

Skill for Claude CodeCodex

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 skills/gbsoss/skill-from-masters/skill-from-github
Any agent
npx skills add GBSOSS/skill-from-masters --skill skill-from-github
Clone the repo
git clone --depth 1 https://github.com/GBSOSS/skill-from-masters

Made for: Claude Code, Codex.

Per session 15 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,151 The whole file, excluding the scripts and references it only reads on demand.
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 $0.00015 $0.01151
Opus 5 $0.00008 $0.00575
Sonnet 5 $0.00003 $0.00230
Haiku 4.5 $0.00002 $0.00115

Measured 2d ago against content hash 3a5c1312f05f, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

skill-from-github 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 2d 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.

skills/skill-from-github/SKILL.md · 191 lines

How it starts

The opening of the file, as written. The whole thing — 191 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Skill from GitHub

When users want to accomplish something, search GitHub for quality projects that solve the problem, understand them deeply, then create a skill based on that knowledge.

When to Use

When users describe a task and you want to find existing tools/projects to learn from:

  • "I want to be able to convert markdown to PDF"
  • "Help me analyze sentiment in customer reviews"
  • "I need to generate API documentation from code"

Workflow

Step 1: Understand User Intent

Clarify what the user wants to achieve:

  • What is the input?
  • What is the expected output?
  • Any constraints (language, framework, etc.)?

Step 2: Search GitHub

Search for projects that solve this problem:

{task keywords} language:{preferred} stars:>100 sort:stars

Search tips:

  • Start broad, then narrow down
  • Try different keyword combinations
  • Include "cli", "tool", "library" if relevant

Quality filters (must meet ALL):

  • Stars > 100 (community validated)
  • Updated within last 12 months (actively maintained)
  • Has README with clear documentation
  • Has actual code (not just awesome-list)

Step 3: Present Options to User

Show top 3-5 candidates:

## Found X projects that can help

### Option 1: [project-name](github-url)
- Stars: xxx | Last updated: xxx
- What it does: one-line description
- Why it's good: specific strength

### Option 2: ...

Which one should I dive into? Or should I search differently?

Wait for user confirmation before proceeding.

Step 4: Deep Dive into Selected Project

Once user selects a project, thoroughly understand it:

  1. Read README - Understand purpose, features, usage
  2. Read core source files - Understand how it works
  3. Check examples - See real usage patterns
  4. Note dependencies - What it relies on
  5. Identify key concepts - The mental model behind it

Extract:

  • Core algorithm/approach
  • Input/output formats
  • Error handling patterns
  • Best practices encoded in the code

Read the full file on GitHub · 191 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. 2d ago First seen · 191 lines · 15 tokens per session scan A 3a5c1312f05f

Subscribe to this mod's changes

skill-from-github is a skill published in the GitHub repository GBSOSS/skill-from-masters (1,576 stars, last pushed 6mo ago), licensed MIT. It adds 15 tokens to every session and 1,151 once invoked, about $0.0001 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.