agent-skills
CLAUDE.md2 starsChanged 5 days agoApache-2.0View on GitHub
A collection of AI agent skills — self-contained Markdown knowledge bases that AI coding assistants (Claude Code, Copilot CLI, Cursor, etc.) load to gain expertise on specific topics. Skills are distributed via npx skills add chaunsin/agent-skills. Each skill lives under skills/<name>/: The description frontmatter field guides skill selection — it describes the capability and scenarios in which an agent should load the skill. Before creating, modifying, or reviewing a skill, read and follow Skill Authoring and Maintenance. It is the…
In this repository12 other files it gives its agents.
- database-designSkill
- drizzle-migration-conflictSkill
- git-stackSkill
- grillSkill
- hugo-to-markdownSkill
- postgresql-cliSkill
- prdSkill
- pre-release-reviewSkill
- rclone-cliSkill
- redis-cliSkill
- to-specSkill
- to-taskSkill
The file
We split this CLAUDE.md into the rules it gives an agent
1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 23
Before creating, modifying, or reviewing a skill, read and follow Skill Authoring and Maintenance.
≈ “Common” is an estimate: at least 1 in 5 of the 26,980 CLAUDE.md files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
CLAUDE.md
This file provides guidance to AI coding assistants when working with code in this repository.
Repository Purpose
A collection of AI agent skills — self-contained Markdown knowledge bases that AI coding assistants (Claude Code, Copilot CLI, Cursor, etc.) load to gain expertise on specific topics. Skills are distributed via npx skills add chaunsin/agent-skills.
Skill Anatomy
Each skill lives under skills/<name>/:
SKILL.md— Entry point with YAML frontmatter (name,description,metadata), essential instructions, and links to conditional detail- Supporting Markdown — Optional detailed syntax, edge cases, workflows, or templates in
references/,reference/, or files alongsideSKILL.md scripts/— (optional) Helper scripts for validation, inventory, or installationassets/— (optional) Templates and other resources used in generated outputagents/— (optional) Agent-specific configurations (e.g., OpenAI agent YAML)
The description frontmatter field guides skill selection — it describes the capability and scenarios in which an agent should load the skill.
Skill Authoring Conventions
Before creating, modifying, or reviewing a skill, read and follow Skill Authoring and Maintenance. It is the canonical guide for content quality, readable structure, standard format and progressive disclosure, evidence for changes, portability, security assessment, and validation.
Current Skills
| Skill | Topic |
database-design | Relational database modeling and schema review — keys, constraints, indexes, transactions, architecture, database-specific DDL, and incremental migrations |
drizzle-migration-conflict | Drizzle Kit migration conflict diagnosis and team workflow guidance - inspect _journal.json, snapshots, migration folders, merge conflicts, drizzle-kit check, and CI prevention |
git-stack | Git workflows — commits, pushes, merge or rebase conflicts, worktree creation and reuse, scoped changes, and reusable validation results |
grill | Decision interviews — examine plans, ideas, and designs through successive questions, map decision dependencies, and clarify assumptions until consensus |
hugo-to-markdown | Hugo documentation and content conversion — inspect local Hugo config, shortcodes, render hooks, front matter, and mounted content roots to produce safe standard Markdown exports |
postgresql-cli | PostgreSQL interactive terminal (psql) — meta-commands, CLI options, formatting, import/export, scripting |
prd | Product requirements — turn ideas and clarified needs into a PRD with users, goals, scope, behavior, and acceptance criteria before implementation |
pre-release-review | Production pre-release code review and deploy-readiness audit — inspect PRs or git ranges for missing migrations, configs, caches, queues, assets, secrets, service order, and rollout risks |
rclone-cli | Rclone cloud storage manager — sync, copy, mount, serve, bisync, crypt, filtering, 70+ providers |
redis-cli | Redis command-line interface — data querying, key scanning, server monitoring, latency analysis, vector search, ACL management, cluster management, scripting |
to-spec | Technical specifications — PRD content by default, with explicitly supplied or selected ADRs, interviews, and Q&A documents supported; source traceability, implementation contracts, and verification criteria |
to-task | Task decomposition — define verifiable work items, priorities, dependencies, and acceptance criteria; save locally or to a user-selected task system without executing or scheduling the work |
Development Commands
This repository contains Markdown documentation and helper scripts only. There is no build system, test suite, or package manager.
Validate Hugo-to-Markdown output
# Detect leftover Hugo syntax and unsafe residue in generated Markdown
python3 skills/hugo-to-markdown/scripts/check_standard_markdown.py --root <markdown-output-dir>
# Inventory Hugo config, shortcodes, render hooks, and content patterns
python3 skills/hugo-to-markdown/scripts/inventory_hugo_rules.py --site-root <hugo-site-dir> --output inventory.jsonInstall a skill locally
# Copy a skill into the Claude Code skills directory
cp -r skills/<skill-name> ~/.claude/skills/Git Conventions
- Conventional commit prefixes:
docs:,refactor:,feat:,init - Chinese commit messages are also accepted — follow the style of recent commits in the branch
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.