Correct, copy-paste starting points. Every snippet here passes agentlint with zero findings — use them as a known-good baseline.
A solid CLAUDE.md
Concise project memory: conventions, commands, and guardrails.
CLAUDE.md
# Project: acme-api
## Stack
- Node 22, TypeScript (strict, ESM), Fastify, Postgres via Prisma.
## Commands
- Install: `npm ci`
- Dev: `npm run dev`
- Test: `npm test` (vitest). Run before every commit.
- Lint/format: `npm run lint` and `npm run format`.
## Conventions
- Prefer pure functions; keep side effects at the edges.
- All new code needs tests. No `any` without a written reason.
- Never commit secrets; use `.env` and `process.env`.
## Out of scope for the agent
- Do not run database migrations against production.
- Do not edit files under `vendor/`.
A reviewer subagent
Valid frontmatter (name, description, scoped tools) plus a body.
.claude/agents/code-reviewer.md
---
name: code-reviewer
description: Reviews staged changes for bugs, security issues, and style. Use proactively before commits.
tools: Read, Grep, Glob, Bash
model: sonnet
---
You are a meticulous senior code reviewer.
When invoked:
1. Run `git diff --staged` to see the changes.
2. Flag correctness bugs, security issues, and missing tests first.
3. Group feedback by severity. Cite file and line. Be concise and specific.
Never modify files yourself — report findings only.
A slash command
Optional frontmatter with allowed-tools and an argument hint.
.claude/commands/changelog.md
---
description: Draft a changelog entry from recent commits.
argument-hint: [version]
allowed-tools: Bash(git log:*), Read
---
Summarize the commits since the last tag into a changelog for version $1.
Recent history:
!`git log --oneline -n 30`
Group entries under Added / Changed / Fixed. Keep it terse.
Portable agent instructions read by Cursor, Copilot, Codex & more — no secrets, no RCE.
AGENTS.md
# AGENTS.md
## Stack
- Go 1.23, chi router, sqlc + Postgres.
## Build & test
- Build: `make build`
- Test: `go test ./...` — run before opening a PR.
## Conventions
- Wrap errors with %w; never panic in request handlers.
- Keep handlers thin; put logic in the service layer.
- Read secrets from the environment — never hardcode them.
A security-auditor subagent
Read-only auditor with scoped tools — finds issues, never edits code.
.claude/agents/security-auditor.md
---
name: security-auditor
description: Audits changes for security issues — injection, authz gaps, unsafe secret handling. Use before merging auth or input-handling code.
tools: Read, Grep, Glob
model: opus
---
You are a security auditor. For the changes in scope:
1. Check input validation, authentication/authorization, and secret handling.
2. Flag injection, SSRF, path traversal, and unsafe deserialization.
3. Report each issue with file:line, a severity, and a concrete fix.
Report findings only — never modify code yourself.
A test-author subagent
Writes focused tests for changed code; runs them and reports.
.claude/agents/test-author.md
---
name: test-author
description: Writes focused unit tests for changed code. Use after implementing a feature or fixing a bug.
tools: Read, Edit, Bash
model: sonnet
---
You write tests, not implementation.
1. Read the changed code and its existing tests.
2. Add tests for the happy path, edge cases, and regressions.
3. Keep tests deterministic and fast; run them and report results.
A /commit slash command
Scoped git tools + a dynamic diff; drafts a conventional-commit message.
.claude/commands/commit.md
---
description: Stage-aware commit — drafts a conventional-commit message from the staged diff.
allowed-tools: Bash(git add:*), Bash(git commit:*), Bash(git status:*), Bash(git diff:*)
---
Review the staged changes and write a concise conventional-commit message, then commit.
Staged summary:
!`git diff --staged --stat`
A PDF-extract skill (SKILL.md)
Valid Agent Skill: name matches the directory, a "Use when…" description, scoped tools.
.claude/skills/pdf-extract/SKILL.md
---
name: pdf-extract
description: Extract text and tables from PDF files and fill PDF forms. Use when the user mentions PDFs, extracting document data, or filling forms.
license: MIT
allowed-tools: Read, Bash(python3:*)
---
# PDF Extract
Extract text and tabular data from PDFs, and fill simple AcroForm fields.
## How to use
1. Confirm the target file is a PDF.
2. Extract text with `pdfplumber`:
```python
import pdfplumber
with pdfplumber.open("document.pdf") as pdf:
text = pdf.pages[0].extract_text()
```
3. For tables, use `page.extract_tables()`.
For form filling and edge cases, see [references/FORMS.md](references/FORMS.md).
A conventional-commits skill (SKILL.md)
Scoped git tools, top-level version/license, a discovery-friendly description.
.claude/skills/conventional-commits/SKILL.md
---
name: conventional-commits
description: Draft Conventional Commits messages from a staged diff. Use when committing changes or when the user asks for a commit message.
license: MIT
version: "1.0.0"
allowed-tools: Bash(git diff:*), Bash(git status:*)
---
# Conventional Commits
Write a single Conventional Commits message for the staged changes.
## Process
1. Read the staged diff: `git diff --staged`.
2. Pick a type: `feat`, `fix`, `docs`, `refactor`, `test`, `chore`.
3. Write `type(scope): summary` in the imperative mood, ≤ 72 chars.
4. Add a body only when the change needs rationale.
Never include secrets or file dumps in the message.
Curated skills (22)
Agent Skills (SKILL.md) synced from upstream repositories and validated by agentlint — every one passes with zero errors. Refreshed automatically each week.
algorithmic-art
Creating algorithmic art using p5.js with seeded randomness and interactive parameter exploration. Use this when users request creating art using code, gener…
Applies Anthropic's official brand colors and typography to any sort of artifact that may benefit from having Anthropic's look-and-feel. Use it when brand co…
Create beautiful visual art in .png and .pdf documents using design philosophy. You should use this skill when the user asks to create a poster, piece of art…
Reference for the Claude API / Anthropic SDK — model ids, pricing, params, streaming, tool use, MCP, agents, caching, token counting, model migration.
TRIGGE…
Guide users through a structured workflow for co-authoring documentation. Use when user wants to write documentation, proposals, technical specs, decision do…
Use this skill whenever the user wants to create, read, edit, or manipulate Word documents (.docx files). Triggers include: any mention of 'Word doc', 'word …
Guidance for distinctive, intentional visual design when building new UI or reshaping an existing one. Helps with aesthetic direction, typography, and making…
A set of resources to help me write all kinds of internal communications, using the formats that my company likes to use. Claude should use this skill whenev…
Guide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use whe…
Use this skill whenever the user wants to do anything with PDF files. This includes reading or extracting text/tables from PDFs, combining or merging multipl…
Use this skill any time a .pptx file is involved in any way — as input, output, or both. This includes: creating slide decks, pitch decks, or presentations; …
Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize a…
Knowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users reques…
Toolkit for styling artifacts with a theme. These artifacts can be slides, docs, reportings, HTML landing pages, etc. There are 10 pre-set themes with colors…
Suite of tools for creating elaborate, multi-component claude.ai HTML artifacts using modern frontend web technologies (React, Tailwind CSS, shadcn/ui). Use …
Toolkit for interacting with and testing local web applications using Playwright. Supports verifying frontend functionality, debugging UI behavior, capturing…
Use this skill any time a spreadsheet file is the primary input or output. This means any task where the user wants to: open, read, edit, or fix an existing …
This skill provides an advanced financial modeling suite with DCF analysis, sensitivity testing, Monte Carlo simulations, and scenario planning for investmen…