Claude Prompt Library

30 Claude Prompts That Write README Files

30 copy-paste prompts

Describe your repo and Claude returns a complete, ready-to-commit README.md as a Markdown artifact โ€” with badges, install, usage, and API sections. Prompts for libraries, web apps, CLIs, APIs, profiles, and templates. Not "give me some text".

In short: This page contains 30 copy-paste ready prompts, organized into 6 categories with a description and pro tip for each. The first 15 prompts are free instantly โ€” no signup needed. Hand-curated and tested by the AI Academy team.

By Louis Corneloup ยท Founder, Techpresso
Last updated ยทHand-curated & tested by the AI Academy team

Library & Package READMEs

5 prompts

npm Package README

1/30

You are an open-source maintainer who writes clear npm package documentation. <context> I am publishing a JavaScript/TypeScript package to npm and need a complete README.md, returned as one ready-to-commit Markdown artifact. </context> <inputs> - Package name and one-liner: [NAME PLUS WHAT IT DOES] - Problem it solves: [PROBLEM] - Install command: [E.G. npm install my-pkg] - Main API / exports: [KEY FUNCTIONS OR CLASSES] - Runtime / peer deps: [NODE VERSION, REACT, ETC] - License and repo: [MIT, GITHUB URL] </inputs> <task> Write a complete README with: an h1 title plus tagline, a shields.io badge row (npm version, downloads, build, license), a one-paragraph description, a Features list, Installation, Quick Start with a runnable example, a Usage / API reference documenting each export with signatures and examples, a Configuration/options table, a TypeScript note, Contributing, and License. Include a table of contents. </task> <constraints> - Valid GitHub-flavored Markdown; fenced code blocks tagged js or ts. - Real, working example code (no lorem); badges use correct shields.io URLs for the package name. - Concise, skimmable, no filler. </constraints> <format> Return the complete README.md as one Markdown artifact, then a short note on which sections to trim for a tiny package. </format>

Generates a complete npm package README with badges, install, and a full API reference as a ready-to-commit artifact.

๐Ÿ’ก

Pro tip: Paste your real exported function signatures so Claude documents each with an accurate example instead of guessing the API.

Python Package (PyPI) README

2/30

You are a Python library author who writes polished PyPI documentation. <context> I am publishing a package to PyPI and need a README that renders well on both GitHub and PyPI, returned as one ready-to-commit Markdown artifact. </context> <inputs> - Package name and purpose: [NAME PLUS WHAT IT DOES] - pip install command: [E.G. pip install mypkg] - Supported Python versions: [E.G. 3.9+] - Core usage example: [MAIN FUNCTION OR CLASS] - Optional extras: [E.G. pip install mypkg[async]] - License and links: [LICENSE, DOCS, REPO] </inputs> <task> Write a README with: title and tagline, badges (PyPI version, Python versions, license, CI), description, Features, Installation including extras, Quickstart with a copy-paste Python example, Usage covering the main API with code, a Configuration section, a link to full docs, Development/testing setup, Contributing, and License. Add a table of contents. </task> <constraints> - Valid Markdown that renders on PyPI (avoid raw HTML PyPI strips); code blocks tagged python and bash. - Real, importable example code; accurate shields.io PyPI badges. - Clear and scannable, no filler. </constraints> <format> Return the complete README.md as one Markdown artifact, then note any GitHub-only syntax to avoid so it renders cleanly on PyPI. </format>

Produces a PyPI-ready Python package README with badges, install extras, and usage examples as a ready-to-commit artifact.

๐Ÿ’ก

Pro tip: Tell Claude which Markdown features PyPI drops so the README renders identically on both GitHub and PyPI.

Go Module README

3/30

You are a Go maintainer who writes idiomatic module documentation. <context> I maintain a Go module and need a README.md that follows Go community conventions, returned as one ready-to-commit Markdown artifact. </context> <inputs> - Module path and purpose: [E.G. github.com/me/mod PLUS WHAT IT DOES] - go get command: [MODULE PATH] - Minimum Go version: [E.G. 1.21] - Key packages / functions: [EXPORTED API] - Concurrency or performance notes: [IF ANY] - License and repo: [LICENSE, URL] </inputs> <task> Write a README with: title and one-liner, badges (Go Reference / pkg.go.dev, Go Report Card, build, license), description, Installation with go get, Usage with an idiomatic runnable example including error handling, an API overview linking to pkg.go.dev, Features, a Benchmarks placeholder, Contributing, and License. Add a table of contents. </task> <constraints> - Valid Markdown; code blocks tagged go; idiomatic Go (handle errors, no panics in examples). - Correct pkg.go.dev and Go Report Card badge URLs for the module path. - Concise, no filler. </constraints> <format> Return the complete README.md as one Markdown artifact, then note how to keep it in sync with godoc comments. </format>

Builds an idiomatic Go module README with pkg.go.dev badges and error-handled examples as a ready-to-commit artifact.

๐Ÿ’ก

Pro tip: Give Claude your real module path so the go get line, import statements, and badge URLs all match exactly.

Rust Crate README

4/30

You are a Rust crate author who writes documentation matching crates.io conventions. <context> I am publishing a crate and need a README.md that doubles as the crates.io description, returned as one ready-to-commit Markdown artifact. </context> <inputs> - Crate name and purpose: [NAME PLUS WHAT IT DOES] - Cargo add / dependency line: [E.G. cargo add mycrate] - Minimum supported Rust version: [MSRV] - Core API example: [MAIN FUNCTION / TRAIT] - Feature flags: [DEFAULT AND OPTIONAL FEATURES] - License and repo: [LICENSE, URL] </inputs> <task> Write a README with: title and tagline, badges (crates.io version, docs.rs, downloads, license, CI), description, Features including a feature-flag table, Installation via Cargo.toml and cargo add, Usage with a compiling example, a link to docs.rs, an MSRV note, Contributing, and a dual-license section. Add a table of contents. </task> <constraints> - Valid Markdown; code blocks tagged rust and toml; the example should compile. - Correct crates.io and docs.rs badge URLs; document feature flags accurately. - Concise and idiomatic, no filler. </constraints> <format> Return the complete README.md as one Markdown artifact, then note how to reuse it as the crate description via the readme field in Cargo.toml. </format>

Generates a crates.io-ready Rust crate README with a feature-flag table, docs.rs badges, and a compiling example as a ready-to-commit artifact.

๐Ÿ’ก

Pro tip: List your real feature flags and Claude will build an accurate feature table plus per-feature usage notes.

Java / Maven Library README

5/30

You are a JVM library maintainer who writes clear Maven and Gradle documentation. <context> I publish a Java library to Maven Central and need a README.md, returned as one ready-to-commit Markdown artifact. </context> <inputs> - Library name and purpose: [NAME PLUS WHAT IT DOES] - Maven coordinates: [GROUPID, ARTIFACTID, VERSION] - Minimum JDK version: [E.G. 17] - Core API example: [MAIN CLASS / METHOD] - Key features: [WHAT IT PROVIDES] - License and repo: [LICENSE, URL] </inputs> <task> Write a README with: title and tagline, badges (Maven Central version, javadoc, build, license), description, Features, Installation snippets for both Maven (pom.xml) and Gradle (build.gradle), Usage with a runnable Java example, a link to Javadoc, a Compatibility/JDK note, Contributing, and License. Add a table of contents. </task> <constraints> - Valid Markdown; code blocks tagged java, xml, and gradle for build files. - Provide both Maven and Gradle install snippets with the real coordinates. - Concise, no filler. </constraints> <format> Return the complete README.md as one Markdown artifact, then note how to auto-update the version badge on each release. </format>

Builds a Maven Central library README with Maven and Gradle install snippets and a Java example as a ready-to-commit artifact.

๐Ÿ’ก

Pro tip: Give Claude your exact groupId and artifactId and it will fill every dependency snippet and the Maven Central badge correctly.

XML tags are just the start. Learn the full Claude workflow.

A growing library of 300+ hands-on AI tutorials covering Claude, ChatGPT, and 50+ tools. New tutorials added every week.

Start 7-Day Free Trial

Web App & SaaS Project READMEs

5 prompts

Full-Stack Web App README

6/30

You are a senior full-stack engineer who writes onboarding-focused project docs. <context> I have a full-stack web app repo and need a README that gets a new developer running locally in minutes, returned as one ready-to-commit Markdown artifact. </context> <inputs> - App name and what it does: [NAME PLUS PURPOSE] - Stack: [E.G. NEXT.JS, POSTGRES, PRISMA] - Prerequisites: [NODE VERSION, DB, PACKAGE MANAGER] - Environment variables: [KEY VARS AND WHAT THEY DO] - Scripts: [E.G. npm run dev, build, test] - Deploy target: [E.G. VERCEL, DOCKER, FLY] </inputs> <task> Write a README with: title and description, a Tech Stack list, Prerequisites, Getting Started (clone, install, copy .env, migrate, run), an Environment Variables table (name, required, description), Available Scripts, a Project Structure tree, Deployment steps, Testing, Contributing, and License. Add a table of contents. </task> <constraints> - Valid Markdown; code blocks tagged bash and env; commands must be copy-paste runnable in order. - The env table lists every variable with required/optional and purpose. - Clear, skimmable, no filler. </constraints> <format> Return the complete README.md as one Markdown artifact, then note which sections to move into a separate CONTRIBUTING.md as the project grows. </format>

Generates a full-stack web app README with an env-var table, ordered setup commands, and deploy steps as a ready-to-commit artifact.

๐Ÿ’ก

Pro tip: Paste your package.json scripts and .env.example so Claude documents the exact commands and every variable with no guessing.

Next.js / React Frontend README

7/30

You are a frontend lead who writes clean documentation for React and Next.js projects. <context> I have a Next.js (or React) frontend and need a README.md, returned as one ready-to-commit Markdown artifact. </context> <inputs> - App name and purpose: [NAME PLUS WHAT IT IS] - Framework and version: [E.G. NEXT.JS 15, VITE + REACT] - Styling / key libraries: [E.G. TAILWIND, SHADCN, REDUX] - Environment / API endpoints: [PUBLIC VARS, API BASE] - Scripts: [DEV, BUILD, LINT, TEST] - Live demo URL: [IF ANY] </inputs> <task> Write a README with: title, a live-demo link and screenshot placeholder, Features, Tech Stack, Getting Started (install and run), Environment Variables, Available Scripts, Folder Structure, Building for production, Deployment, and License. Add a table of contents and a badge row (framework, license, deploy status). </task> <constraints> - Valid Markdown; code blocks tagged bash; include a screenshot placeholder with descriptive alt text. - Accurate NEXT_PUBLIC / VITE env var naming; no filler. - Skimmable with clear headings. </constraints> <format> Return the complete README.md as one Markdown artifact, then note where to add a demo GIF and how to link the deployed preview. </format>

Produces a Next.js/React frontend README with a demo link, screenshot slot, scripts, and structure as a ready-to-commit artifact.

๐Ÿ’ก

Pro tip: Tell Claude your deploy URL and it will wire a live-demo badge and a Deploy button placeholder into the header.

SaaS README With Architecture Overview

8/30

You are a staff engineer who documents SaaS codebases for new team members. <context> I need a README for a SaaS product repo that explains architecture and onboarding, returned as one ready-to-commit Markdown artifact. </context> <inputs> - Product name and what it does: [NAME PLUS PURPOSE] - Architecture: [SERVICES, DB, QUEUE, CACHE, THIRD-PARTY] - Stack per layer: [FRONTEND, BACKEND, INFRA] - Local setup steps: [HOW TO RUN] - Env vars / secrets handling: [WHERE SECRETS LIVE] - Team conventions: [BRANCHING, PR, DEPLOY] </inputs> <task> Write a README with: overview, an Architecture section with a Mermaid diagram of services and data flow, Tech Stack by layer, Local Development setup, an Environment/secrets note, Available Scripts, Deployment/CI overview, Team Conventions (branching, PRs, releases), Troubleshooting, and links to deeper docs. Add a table of contents. </task> <constraints> - Valid Markdown; include a fenced mermaid diagram that GitHub renders; code blocks tagged bash. - Do not print real secrets; describe where they live and how to obtain them. - Clear and thorough but skimmable. </constraints> <format> Return the complete README.md as one Markdown artifact, then note which parts belong in an internal wiki versus the public README. </format>

Builds a SaaS repo README with a Mermaid architecture diagram, a layered stack, and team conventions as a ready-to-commit artifact.

๐Ÿ’ก

Pro tip: Describe your services and their connections and Claude will render them as a GitHub-native Mermaid diagram instead of an external image.

Self-Hosted Docker Compose App README

9/30

You are a DevOps engineer who writes self-hosting documentation. <context> I ship a self-hostable app via Docker Compose and need a README that helps users deploy it, returned as one ready-to-commit Markdown artifact. </context> <inputs> - App name and purpose: [NAME PLUS WHAT IT DOES] - Services in the stack: [APP, DB, REVERSE PROXY, ETC] - Required env / config: [KEY SETTINGS] - Ports and volumes: [PORTS, PERSISTENT DATA PATHS] - Update / backup process: [HOW TO UPGRADE AND BACK UP] - License and repo: [LICENSE, URL] </inputs> <task> Write a README with: title and description, badges (docker pulls, image size, license), Features, Requirements (Docker, Compose versions), Quick Start with a copy-paste docker-compose.yml and .env, a Configuration table, Ports and Volumes, a reverse-proxy/HTTPS note, Updating, Backup and Restore, Troubleshooting, and License. Add a table of contents. </task> <constraints> - Valid Markdown; include a runnable docker-compose.yml fenced as yaml and an .env example. - Document every env var, port, and volume; commit no real secrets. - Practical and complete, no filler. </constraints> <format> Return the complete README.md as one Markdown artifact, then note how to add a one-line deploy script or a Coolify/CapRover template. </format>

Generates a self-hosted Docker Compose app README with a runnable compose file, ports/volumes, and backup steps as a ready-to-commit artifact.

๐Ÿ’ก

Pro tip: List your real services, ports, and volumes and Claude will produce a compose file plus a matching config table users can trust.

Contributor-Focused Open-Source README

10/30

You are an open-source community manager who writes contributor-friendly READMEs. <context> I want to grow contributions to my open-source web project and need a welcoming, well-structured README, returned as one ready-to-commit Markdown artifact. </context> <inputs> - Project name and mission: [NAME PLUS WHY IT EXISTS] - What it does / key features: [FEATURES] - Tech stack: [STACK] - How to run locally: [SETUP STEPS] - Ways to contribute: [CODE, DOCS, ISSUES, TRANSLATIONS] - Community links: [DISCORD, DISCUSSIONS, CODE OF CONDUCT] </inputs> <task> Write a README with: an eye-catching header (logo placeholder, tagline, badges for stars, license, PRs welcome, build), a clear description, Features, a Demo/screenshot slot, Getting Started, How to Contribute (good-first-issue pointer, dev setup, PR process), a Roadmap, Community and Support links, a Contributors placeholder, a Code of Conduct link, and License. Add a table of contents. </task> <constraints> - Valid Markdown; welcoming and inclusive tone; code blocks tagged bash. - Include a PRs-welcome badge and a star badge; link to CONTRIBUTING.md and CODE_OF_CONDUCT.md. - Motivating but concrete, no filler. </constraints> <format> Return the complete README.md as one Markdown artifact, then note which community files (CONTRIBUTING, CODE_OF_CONDUCT, issue templates) to create alongside it. </format>

Builds a contributor-focused open-source README with badges, a roadmap, and a clear contribution path as a ready-to-commit artifact.

๐Ÿ’ก

Pro tip: Point Claude at your good-first-issue label so it writes an exact your-first-contribution walkthrough newcomers can follow.

CLI Tool READMEs

5 prompts

CLI Tool README With Command Reference

11/30

You are a developer-tools author who writes precise CLI documentation. <context> I built a command-line tool and need a README with a full command reference, returned as one ready-to-commit Markdown artifact. </context> <inputs> - Tool name and purpose: [NAME PLUS WHAT IT DOES] - Install command: [E.G. npm i -g, brew install, cargo install] - Commands and subcommands: [LIST WITH WHAT EACH DOES] - Common flags/options: [GLOBAL AND PER-COMMAND FLAGS] - Example workflows: [TYPICAL USAGE] - License and repo: [LICENSE, URL] </inputs> <task> Write a README with: title and tagline, badges (version, downloads, license), a description, an animated-demo placeholder, Installation, Usage with a Commands reference table (command, description, example), detailed sections per major command with flags, Configuration, real-workflow Examples, Contributing, and License. Add a table of contents. </task> <constraints> - Valid Markdown; code blocks tagged bash showing real invocations and sample output. - Every command and flag documented in a table plus an example; no filler. - Skimmable with copy-paste-ready commands. </constraints> <format> Return the complete README.md as one Markdown artifact, then note how to keep the command reference in sync with your --help output. </format>

Generates a CLI tool README with a full command/flag reference table and workflow examples as a ready-to-commit artifact.

๐Ÿ’ก

Pro tip: Paste your tool's --help output and Claude will turn it into an accurate, formatted command reference table.

Homebrew-Installable CLI README

12/30

You are a macOS and Linux CLI author who distributes via Homebrew. <context> I distribute a CLI through a Homebrew tap and need a README, returned as one ready-to-commit Markdown artifact. </context> <inputs> - Tool name and purpose: [NAME PLUS WHAT IT DOES] - Tap and formula: [E.G. brew tap me/tools; brew install me/tools/tool] - Alternative installs: [BINARY DOWNLOAD, GO/CARGO INSTALL] - Core commands: [MAIN USAGE] - Config file location: [E.G. ~/.config/tool] - License and repo: [LICENSE, URL] </inputs> <task> Write a README with: title and tagline, badges (release, license, Homebrew), description, Installation covering the Homebrew tap plus alternatives, Quick Start, Usage with examples, Configuration (file path and options), Updating and Uninstalling via brew, Shell completion setup, Contributing, and License. Add a table of contents. </task> <constraints> - Valid Markdown; code blocks tagged bash; brew commands must be exact for the tap and formula. - Cover update and uninstall paths, not just install; no filler. - Concise and practical. </constraints> <format> Return the complete README.md as one Markdown artifact, then note how to document the tap-then-install two-step so users do not miss it. </format>

Produces a Homebrew-distributed CLI README with tap install, update/uninstall, and shell completion steps as a ready-to-commit artifact.

๐Ÿ’ก

Pro tip: Give Claude your exact tap name and formula so every brew command in the README works when copied verbatim.

Global npm / npx CLI README

13/30

You are a Node.js CLI author who ships tools via npm and npx. <context> I publish a Node CLI to npm meant to be run globally or via npx, and need a README, returned as one ready-to-commit Markdown artifact. </context> <inputs> - Package and command name: [NPM NAME AND BINARY NAME] - What it does: [PURPOSE] - npx one-liner: [E.G. npx create-thing my-app] - Global install: [npm i -g name] - Options / prompts: [FLAGS AND INTERACTIVE PROMPTS] - Node version required: [E.G. 18+] </inputs> <task> Write a README with: title and tagline, badges (npm version, downloads, node version, license), description, a Use-it-in-one-command npx section, Global Installation, Usage with flags and an example session, an Options table, an interactive-prompts walkthrough if applicable, Requirements, Contributing, and License. Add a table of contents. </task> <constraints> - Valid Markdown; code blocks tagged bash; show the npx one-liner prominently near the top. - Accurate npm and node badges; document every flag; no filler. - Skimmable and copy-paste friendly. </constraints> <format> Return the complete README.md as one Markdown artifact, then note whether to lead with npx or global install for your audience. </format>

Builds a Node npm/npx CLI README that leads with a one-command npx usage and a full options table as a ready-to-commit artifact.

๐Ÿ’ก

Pro tip: Tell Claude if your CLI is a scaffolder (like create-*) so it foregrounds the npx one-liner the way users expect.

Shell Script / Dotfiles README

14/30

You are a systems tinkerer who documents shell scripts and dotfiles clearly. <context> I have a shell script or dotfiles repo and need a README so others (and future me) can install and use it safely, returned as one ready-to-commit Markdown artifact. </context> <inputs> - Repo name and purpose: [NAME PLUS WHAT IT SETS UP] - Install method: [E.G. curl | bash, git clone + ./install.sh, stow] - What it changes: [FILES, SHELL, TOOLS INSTALLED] - Supported OS/shells: [MACOS, LINUX, ZSH, BASH] - Key scripts/aliases: [WHAT THEY DO] - License: [LICENSE] </inputs> <task> Write a README with: title and description, a Requirements/Supported-systems note, Installation with a clear warning about piping to a shell and how to review the script first, What Gets Installed or Changed, Usage of key scripts and aliases, Customization, Uninstall/Revert steps, a Safety note, and License. Add a table of contents. </task> <constraints> - Valid Markdown; code blocks tagged bash; include an explicit review-before-running caution for any curl-piped-to-shell install. - Document what the script modifies and how to undo it; no filler. - Practical and honest about side effects. </constraints> <format> Return the complete README.md as one Markdown artifact, then note the safety disclaimers worth keeping for any install-script repo. </format>

Generates a shell-script/dotfiles README with install, what-it-changes, uninstall, and a safety caution as a ready-to-commit artifact.

๐Ÿ’ก

Pro tip: List every file and setting your script touches so Claude writes an accurate what-gets-changed section and a real revert path.

Config-File + Subcommands CLI README

15/30

You are a developer-experience writer documenting a configurable multi-command CLI. <context> My CLI has multiple subcommands and a config file, and I need a README that covers both, returned as one ready-to-commit Markdown artifact. </context> <inputs> - Tool name and purpose: [NAME PLUS WHAT IT DOES] - Install: [HOW TO INSTALL] - Subcommands: [LIST WITH DESCRIPTIONS] - Config file: [FORMAT, LOCATION, KEY OPTIONS] - Environment overrides: [ENV VARS THAT OVERRIDE CONFIG] - License and repo: [LICENSE, URL] </inputs> <task> Write a README with: title and tagline, badges, description, Installation, Quick Start, a Commands section with a subcommand table and per-command examples, a Configuration section showing a sample config file with commented options, a Config Precedence note (flags > env > file > defaults), Examples, Contributing, and License. Add a table of contents. </task> <constraints> - Valid Markdown; code blocks tagged bash plus the config format (yaml/toml/json); show a commented sample config. - Clearly state option precedence; document every subcommand; no filler. - Skimmable and precise. </constraints> <format> Return the complete README.md as one Markdown artifact, then note how to generate the config-options table from your schema. </format>

Builds a multi-subcommand CLI README with a commented sample config, precedence rules, and per-command examples as a ready-to-commit artifact.

๐Ÿ’ก

Pro tip: Give Claude your config schema and it will produce an options table plus a fully commented sample config file.

API & Service READMEs

5 prompts

REST API README

16/30

You are a backend engineer who writes clear REST API documentation. <context> I have a REST API and need a README documenting auth, endpoints, and examples, returned as one ready-to-commit Markdown artifact. </context> <inputs> - API name and purpose: [NAME PLUS WHAT IT SERVES] - Base URL and versioning: [E.G. https://api.x.com/v1] - Auth method: [API KEY, BEARER, OAUTH] - Key endpoints: [METHOD, PATH, WHAT EACH DOES] - Rate limits / errors: [LIMITS AND ERROR FORMAT] - Run locally: [SETUP STEPS] </inputs> <task> Write a README with: title and description, badges (build, license, API version), Overview, Base URL and Authentication with a header example, an Endpoints reference table (method, path, description), detailed examples per key endpoint with a curl request and JSON response, an Errors section with the standard error shape and status codes, Rate Limiting, Local Development, and License. Add a table of contents. </task> <constraints> - Valid Markdown; code blocks tagged bash for curl and json for payloads; realistic request/response examples. - Never print real API keys; use a placeholder token; keep the error shape consistent. - Precise and complete, no filler. </constraints> <format> Return the complete README.md as one Markdown artifact, then note when to graduate from a README to OpenAPI/Swagger. </format>

Generates a REST API README with an endpoints table, curl/JSON examples, auth, and an error reference as a ready-to-commit artifact.

๐Ÿ’ก

Pro tip: Paste two or three real endpoints with sample responses and Claude will match the JSON shape and status codes across the whole doc.

GraphQL API README

17/30

You are a GraphQL API author who documents schemas and operations clearly. <context> I have a GraphQL API and need a README covering the schema, queries, mutations, and auth, returned as one ready-to-commit Markdown artifact. </context> <inputs> - API name and purpose: [NAME PLUS WHAT IT DOES] - Endpoint URL: [GRAPHQL ENDPOINT] - Auth: [HOW CLIENTS AUTHENTICATE] - Key types: [MAIN OBJECT TYPES] - Key queries and mutations: [WITH WHAT THEY RETURN] - Playground / introspection: [AVAILABLE OR NOT] </inputs> <task> Write a README with: title and description, badges, Overview, Endpoint and Authentication, a Getting Started query example, a Schema Overview of key types, a Queries section with an example query plus variables plus response, a Mutations section similarly, Error handling, Rate limits, a Playground link, Local Development, and License. Add a table of contents. </task> <constraints> - Valid Markdown; code blocks tagged graphql for operations and json for responses and variables. - Show real, executable example operations with variables; use a placeholder auth token. - Clear and complete, no filler. </constraints> <format> Return the complete README.md as one Markdown artifact, then note how to auto-generate schema docs and keep the README examples in sync. </format>

Produces a GraphQL API README with a schema overview, example queries/mutations, and variables as a ready-to-commit artifact.

๐Ÿ’ก

Pro tip: Paste your SDL for the key types and Claude will write example operations whose fields exactly match your schema.

Microservice README (Health + Observability)

18/30

You are a platform engineer who documents microservices for on-call and onboarding. <context> I own a microservice and need a README covering how to run it, its env, health, and observability, returned as one ready-to-commit Markdown artifact. </context> <inputs> - Service name and responsibility: [NAME PLUS WHAT IT OWNS] - Stack and dependencies: [LANGUAGE, DB, QUEUE, UPSTREAM/DOWNSTREAM SERVICES] - Env vars: [KEY CONFIG] - Endpoints/consumers: [WHO CALLS IT, WHAT IT EXPOSES] - Health/metrics endpoints: [E.G. /healthz, /metrics] - Deploy/runbook basics: [HOW IT DEPLOYS, ALERTS] </inputs> <task> Write a README with: title and responsibility statement, Architecture context (upstream/downstream with a small Mermaid diagram), Tech Stack, Local Development (run plus dependencies via Docker), an Environment Variables table, the API/interface it exposes, Health and Observability (health, readiness, metrics, logs, tracing), Deployment, an on-call Runbook section (common failures and fixes), and links. Add a table of contents. </task> <constraints> - Valid Markdown; include a fenced mermaid diagram of dependencies; code blocks tagged bash. - Document health/readiness/metrics endpoints and key alerts; include no real secrets. - Operationally useful, no filler. </constraints> <format> Return the complete README.md as one Markdown artifact, then note which runbook items belong in a dedicated RUNBOOK.md. </format>

Builds a microservice README with a dependency diagram, env table, health/observability, and an on-call runbook as a ready-to-commit artifact.

๐Ÿ’ก

Pro tip: List your upstream and downstream services and Claude will draw the dependency Mermaid diagram and tailor the runbook to those failure modes.

Webhook / Integration Service README

19/30

You are an integrations engineer who documents webhook-based services. <context> I built a service that sends or receives webhooks and need a README explaining setup, payloads, and verification, returned as one ready-to-commit Markdown artifact. </context> <inputs> - Service name and purpose: [NAME PLUS WHAT IT INTEGRATES] - Direction: [RECEIVES WEBHOOKS / SENDS WEBHOOKS / BOTH] - Events and payloads: [EVENT TYPES AND SAMPLE PAYLOAD] - Signature/verification: [HMAC HEADER, SECRET HANDLING] - Setup/config: [HOW TO REGISTER THE WEBHOOK] - Retry behavior: [RETRIES, TIMEOUTS] </inputs> <task> Write a README with: title and description, an Overview of the integration, Setup (registering the endpoint, configuring the secret), an Events reference table (event, when it fires, payload), sample JSON payloads per event, Signature Verification with a code example, Handling retries and idempotency, Testing (how to send a test event), Troubleshooting, and License. Add a table of contents. </task> <constraints> - Valid Markdown; code blocks tagged json for payloads and a language for the verification snippet; use a placeholder signing secret. - Explain idempotency and retry handling; document every event; no filler. - Precise and security-aware. </constraints> <format> Return the complete README.md as one Markdown artifact, then note the verification pitfalls (raw body, timing-safe compare) worth calling out. </format>

Generates a webhook/integration service README with event payloads, signature verification, and retry/idempotency guidance as a ready-to-commit artifact.

๐Ÿ’ก

Pro tip: Paste one real webhook payload and your signing scheme so Claude documents the exact fields and a correct verification example.

API Client SDK README

20/30

You are an SDK author who writes onboarding docs for API client libraries. <context> I built a client SDK that wraps an API and need a README, returned as one ready-to-commit Markdown artifact. </context> <inputs> - SDK name and language: [NAME PLUS LANGUAGE] - Install command: [PACKAGE INSTALL] - Auth/init: [HOW TO CONSTRUCT THE CLIENT WITH A KEY] - Key methods/resources: [MAIN CALLS AND WHAT THEY DO] - Async / pagination behavior: [IF ANY] - License and repo: [LICENSE, URL] </inputs> <task> Write a README with: title and tagline, badges (package version, license, docs), description, Installation, Authentication and client initialization, Quick Start with a real call and response, a Usage section covering key resources/methods with examples, Error handling, Pagination/async patterns, Configuration (timeouts, retries), Contributing, and License. Add a table of contents. </task> <constraints> - Valid Markdown; code blocks tagged for the SDK's language; read the API key from an env var placeholder. - Show idiomatic init, a real call, and error handling; document key methods; no filler. - Clear and example-driven. </constraints> <format> Return the complete README.md as one Markdown artifact, then note how to keep the examples tested so they never drift from the SDK. </format>

Builds an API client SDK README with init, authenticated calls, error handling, and pagination examples as a ready-to-commit artifact.

๐Ÿ’ก

Pro tip: Tell Claude your SDK's language and top three methods so it writes idiomatic, copy-paste examples that match your surface.

Profile & Portfolio READMEs

5 prompts

GitHub Profile README

21/30

You are a developer-brand designer who crafts standout GitHub profile READMEs. <context> I want a GitHub profile README (the special username/username repo shown on my profile) returned as one ready-to-commit Markdown artifact. </context> <inputs> - Name and role: [NAME PLUS WHAT YOU DO] - Focus / current work: [WHAT YOU ARE BUILDING OR LEARNING] - Top skills / tech: [LANGUAGES, TOOLS] - Notable projects: [2-4 REPOS WITH ONE-LINERS] - Links: [WEBSITE, LINKEDIN, TWITTER/X, EMAIL] - Vibe: [MINIMAL, PLAYFUL, PROFESSIONAL] </inputs> <task> Write a profile README with: a friendly header intro, an About Me blurb, a What I am working on / learning section, a Tech Stack with shields.io skill badges, Featured Projects with links and one-line descriptions, a GitHub stats widget block (github-readme-stats), a top-languages card, a social/contact badge row, and an optional fun fact or quote. Match the requested vibe. </task> <constraints> - Valid Markdown; use shields.io badges and standard github-readme-stats image URLs with a [USERNAME] placeholder. - Keep it skimmable and personality-forward; no filler; center-align the header tastefully. - Set image alt text for accessibility. </constraints> <format> Return the complete README.md as one Markdown artifact, then note which dynamic widgets need a username swap and any that require a token. </format>

Generates a standout GitHub profile README with skill badges, stats widgets, and featured projects as a ready-to-commit artifact.

๐Ÿ’ก

Pro tip: Give Claude your exact GitHub username so every stats widget and badge URL renders with your real data on first commit.

Portfolio Project README With Screenshots

22/30

You are a designer-developer who documents portfolio projects to impress reviewers. <context> I want a README for a portfolio/showcase project that recruiters and peers will read, returned as one ready-to-commit Markdown artifact. </context> <inputs> - Project name and what it is: [NAME PLUS PURPOSE] - Why you built it / your role: [MOTIVATION, ROLE] - Tech stack: [STACK] - Live demo and repo: [URLS] - Key features and challenges solved: [HIGHLIGHTS] - Screenshots available: [WHAT VIEWS] </inputs> <task> Write a README with: a header (title, one-line pitch, live-demo and repo badges), a hero screenshot placeholder, Overview (problem, solution, your role), Features, a Screenshots/Demo gallery with placeholders and captions, a Tech Stack with badges, a How It Is Built architecture note, Getting Started, Challenges and Learnings, and License. Add a table of contents. </task> <constraints> - Valid Markdown; screenshot placeholders with descriptive alt text and captions; badges for stack, demo, license. - Emphasize outcomes and decisions, not just features; no filler. - Polished and recruiter-friendly. </constraints> <format> Return the complete README.md as one Markdown artifact, then note where a short demo GIF has the most impact for reviewers. </format>

Produces a portfolio project README with a screenshot gallery, a how-it-is-built note, and challenges/learnings as a ready-to-commit artifact.

๐Ÿ’ก

Pro tip: Tell Claude the one technical challenge you are proudest of solving so it foregrounds that in the Challenges and Learnings section.

Awesome List README

23/30

You are a curator who builds well-organized awesome-list READMEs. <context> I am creating an awesome-list repo (curated resources on a topic) and need a README that follows awesome-list conventions, returned as one ready-to-commit Markdown artifact. </context> <inputs> - Topic: [WHAT THE LIST CURATES] - Categories: [SECTIONS TO ORGANIZE BY] - Seed entries: [KNOWN ITEMS WITH LINKS AND DESCRIPTIONS] - Audience: [WHO IT IS FOR] - Contribution rules: [WHAT QUALIFIES] </inputs> <task> Write an awesome-list README with: a title and the awesome badge, a one-line description, a Contents table of contents, well-organized category sections each with entries formatted as [Name](url) - Short description., a Contributing section pointing to contribution guidelines, a Related Lists section, and a License (typically CC0). Populate categories with the seed entries and sensible structure. </task> <constraints> - Valid Markdown following the awesome-list format exactly (entry pattern, logical order, the awesome badge). - Descriptions are one concise sentence, no marketing fluff; every entry is a real Markdown link with a [PLACEHOLDER] URL where unknown. - Clean and consistent. </constraints> <format> Return the complete README.md as one Markdown artifact, then note the awesome-lint rules to pass before submitting to the awesome index. </format>

Builds an awesome-list README with the standard badge, categorized entries, and contribution rules as a ready-to-commit artifact.

๐Ÿ’ก

Pro tip: Give Claude your category structure up front so entries land in the right sections and the table of contents matches exactly.

Widget-Rich Profile README

24/30

You are a developer-branding specialist focused on dynamic README widgets. <context> I want a data-rich personal profile README packed with dynamic badges and stats widgets, returned as one ready-to-commit Markdown artifact. </context> <inputs> - GitHub username: [USERNAME] - Headline: [ONE-LINE WHO YOU ARE] - Skills to badge: [LANGUAGES, FRAMEWORKS, TOOLS] - Widgets wanted: [STATS, STREAK, TOP LANGS, TROPHIES, VISITOR COUNT] - Socials: [LINKS TO BADGE] - Theme: [E.G. DARK, RADICAL, TOKYONIGHT] </inputs> <task> Write a profile README with: a centered header with the headline and a typing-SVG or banner placeholder, a social badges row, a Connect With Me block, a Tech Stack badge wall grouped by category, a Dynamic Stats section (github-readme-stats card, streak stats, top languages, optional trophies and visitor counter) all themed consistently, and a footer. Apply the chosen theme across every widget. </task> <constraints> - Valid Markdown; use correct widget image URLs (github-readme-stats, streak-stats, github-profile-trophy) with [USERNAME] and matching theme params. - Apply one consistent theme; set image alt text; no filler. - Visually balanced and centered where appropriate. </constraints> <format> Return the complete README.md as one Markdown artifact, then list which widgets need a deployed instance or token and which work out of the box. </format>

Generates a widget-rich profile README with themed stats, streak, top-languages, and social badges as a ready-to-commit artifact.

๐Ÿ’ก

Pro tip: Pick one theme name and Claude will apply it to every widget URL so the whole profile looks cohesive instead of mismatched.

GitHub Organization Profile README

25/30

You are a brand and community lead who writes GitHub organization profile READMEs. <context> I manage a GitHub organization and want a profile README (the .github repo's profile/README.md shown on the org page), returned as one ready-to-commit Markdown artifact. </context> <inputs> - Organization name and mission: [NAME PLUS WHAT IT DOES] - What you build / flagship products: [KEY REPOS OR PRODUCTS] - Who you are for: [USERS / CONTRIBUTORS] - Ways to get involved: [CONTRIBUTE, JOBS, COMMUNITY] - Links: [WEBSITE, DOCS, SOCIALS] - Vibe: [PROFESSIONAL, COMMUNITY-DRIVEN, ETC] </inputs> <task> Write an org profile README with: a centered header (logo placeholder, tagline, badges), a mission statement, a What We Build grid of flagship repos/products with one-liners and links, a Get Involved section (contribute, join, hire), a Community and Support block, org highlights or stats, a social links row, and a footer. Match the requested vibe. </task> <constraints> - Valid Markdown; logo placeholder with alt text; shields.io badges for key links. - Warm, on-brand, and scannable; link real repos with [PLACEHOLDER] URLs where unknown; no filler. - Balanced layout suitable for an org landing page. </constraints> <format> Return the complete README.md as one Markdown artifact, then note that it must live in the org's .github repo at profile/README.md to show on the org page. </format>

Builds a GitHub organization profile README with a flagship-repo grid and a get-involved section as a ready-to-commit artifact.

๐Ÿ’ก

Pro tip: Have Claude's closing note remind you to place the file at profile/README.md in the .github repo, or it will not render on the org page.

Most people use 10% of Claude. Tutorials unlock the rest.

AI Academy: 300+ hands-on tutorials on Claude, ChatGPT, Midjourney, and 50+ AI tools. New tutorials added every week.

Start Your Free Trial

Template & Boilerplate READMEs

5 prompts

Starter Template / Boilerplate README

26/30

You are a boilerplate author who writes clear starter-kit documentation. <context> I built a starter template/boilerplate others will clone to bootstrap projects and need a README, returned as one ready-to-commit Markdown artifact. </context> <inputs> - Template name and purpose: [NAME PLUS WHAT IT BOOTSTRAPS] - Stack and included tooling: [FRAMEWORK, LINTING, TESTING, CI] - How to use it: [USE-THIS-TEMPLATE / degit / clone] - What is included: [FOLDERS, PRECONFIGURED FEATURES] - First steps after cloning: [RENAME, ENV, INSTALL, RUN] - License and repo: [LICENSE, URL] </inputs> <task> Write a README with: title and tagline, badges (template, license, CI), a What's Included feature checklist, a Use This Template section (GitHub template button, degit, or clone), Getting Started (install, configure, run), a Project Structure tree, a Make It Yours checklist (rename, update metadata, replace placeholders), Included Scripts, Deployment, and License. Add a table of contents. </task> <constraints> - Valid Markdown; code blocks tagged bash; include the exact use-this-template and degit commands. - Provide a post-clone checklist so users know what to change first; no filler. - Skimmable and action-oriented. </constraints> <format> Return the complete README.md as one Markdown artifact, then note which lines new users forget to change and how to flag them. </format>

Generates a starter template README with a what's-included checklist, use-this-template steps, and a make-it-yours checklist as a ready-to-commit artifact.

๐Ÿ’ก

Pro tip: List every placeholder in your template (name, author, repo URL) so Claude builds a post-clone checklist that catches all of them.

Monorepo README

27/30

You are a build engineer who documents monorepos for contributors. <context> I maintain a monorepo with multiple packages/apps and need a README explaining the layout and workflows, returned as one ready-to-commit Markdown artifact. </context> <inputs> - Repo name and purpose: [NAME PLUS WHAT IT CONTAINS] - Tooling: [E.G. TURBOREPO, NX, PNPM WORKSPACES] - Packages/apps: [LIST WITH WHAT EACH IS] - Common tasks: [BUILD, DEV, TEST, LINT ACROSS PACKAGES] - Versioning/release: [E.G. CHANGESETS] - License: [LICENSE] </inputs> <task> Write a README with: title and overview, an Apps and Packages table (path, name, description), a Repository Structure tree, Prerequisites, Getting Started (install at root), Common Commands (run across all, filter to one package), how to add a new package, Versioning and Releases, a CI overview, Contributing, and License. Add a table of contents. </task> <constraints> - Valid Markdown; code blocks tagged bash showing workspace-aware commands (filtering to a package). - Accurately reflect the chosen monorepo tool's command syntax; document every package; no filler. - Clear structure and scannable. </constraints> <format> Return the complete README.md as one Markdown artifact, then note which per-package details belong in each package's own README versus the root. </format>

Builds a monorepo README with a packages table, workspace-aware commands, and a release workflow as a ready-to-commit artifact.

๐Ÿ’ก

Pro tip: Name your monorepo tool (Turborepo, Nx, pnpm) so Claude uses its exact filter and run syntax in every command example.

GitHub Template Repo Getting-Started README

28/30

You are a developer-onboarding writer for reusable GitHub template repositories. <context> I created a GitHub template repository and want a README that guides first-time users the moment they generate a repo from it, returned as one ready-to-commit Markdown artifact. </context> <inputs> - Template name and what it produces: [NAME PLUS OUTCOME] - Target user: [WHO USES IT] - Setup steps after generating: [ORDERED STEPS] - Configurable placeholders: [WHAT TO REPLACE] - Included CI/automation: [WORKFLOWS, HOOKS] - License: [LICENSE] </inputs> <task> Write a README with: a header noting Generated From a Template - Start Here, a numbered First Do These Steps checklist (replace placeholders, set secrets, update this README), What's Included, Configuration, Included Automation (CI, dependabot, etc.), how to keep it updated, a Delete This Section Once Set Up note, and License. Add a table of contents. </task> <constraints> - Valid Markdown; code blocks tagged bash; the first-run checklist is unmissable and ordered. - Include a clearly marked block for users to delete after setup; no filler. - Action-first and beginner-friendly. </constraints> <format> Return the complete README.md as one Markdown artifact, then note how to structure the delete-after-setup section so it is obvious but not embarrassing if left in. </format>

Produces a template-repo getting-started README with an ordered first-run checklist and a delete-after-setup block as a ready-to-commit artifact.

๐Ÿ’ก

Pro tip: Ask Claude to mark the throwaway onboarding block with an HTML-comment banner so users clearly see what to remove once configured.

Terraform Module / IaC README

29/30

You are a cloud infrastructure engineer who documents Terraform modules. <context> I publish a reusable Terraform module and need a README that follows Terraform Registry conventions, returned as one ready-to-commit Markdown artifact. </context> <inputs> - Module name and what it provisions: [NAME PLUS RESOURCES] - Provider and versions: [AWS/GCP/AZURE, TERRAFORM VERSION] - Inputs (variables): [KEY VARIABLES, TYPES, DEFAULTS] - Outputs: [WHAT IT RETURNS] - Example usage: [A TYPICAL MODULE BLOCK] - License and repo: [LICENSE, URL] </inputs> <task> Write a README with: title and description, badges (Terraform Registry, license), a Usage section with a copy-paste module block, a Requirements table (Terraform and provider versions), a Providers table, an Inputs table (name, description, type, default, required), an Outputs table, one or more Examples, Notes/caveats, and License. Add a table of contents. </task> <constraints> - Valid Markdown; code blocks tagged hcl; Inputs/Outputs/Requirements formatted as terraform-docs-style tables. - Document every variable and output with type and default; use a realistic example block; no filler. - Registry-compatible structure. </constraints> <format> Return the complete README.md as one Markdown artifact, then note how terraform-docs can auto-generate the Inputs and Outputs tables from your code. </format>

Builds a Terraform module README with terraform-docs-style inputs/outputs tables and a copy-paste usage block as a ready-to-commit artifact.

๐Ÿ’ก

Pro tip: Paste your variables.tf and outputs.tf and Claude will fill the Inputs and Outputs tables with exact types and defaults.

GitHub Action / Reusable Workflow README

30/30

You are a CI/CD engineer who documents GitHub Actions for the marketplace. <context> I built a GitHub Action (or reusable workflow) and need a README suitable for the GitHub Marketplace, returned as one ready-to-commit Markdown artifact. </context> <inputs> - Action name and what it does: [NAME PLUS PURPOSE] - Inputs: [INPUT NAMES, DESCRIPTIONS, DEFAULTS, REQUIRED] - Outputs: [OUTPUT NAMES AND MEANINGS] - Example workflow usage: [A STEP THAT USES IT] - Runs on / requirements: [PERMISSIONS, RUNNER, SECRETS] - License and repo: [LICENSE, URL] </inputs> <task> Write a README with: title and tagline, badges (marketplace, release, license), description, a Usage section with a complete example workflow YAML that uses the action, an Inputs table (name, description, required, default), an Outputs table, required Permissions/secrets, Advanced examples (matrix, conditional), a Versioning note (pin to a major tag), Contributing, and License. Add a table of contents. </task> <constraints> - Valid Markdown; the example workflow is valid, copy-paste YAML fenced as yaml; reference the action as owner/repo@v1. - Document every input and output; note required permissions and secrets; no filler. - Marketplace-ready and precise. </constraints> <format> Return the complete README.md as one Markdown artifact, then note the branding and versioning steps needed to list it on the GitHub Marketplace. </format>

Generates a GitHub Action README with a valid example workflow, inputs/outputs tables, and permissions notes as a ready-to-commit artifact.

๐Ÿ’ก

Pro tip: Paste your action.yml and Claude will turn its inputs and outputs into accurate tables and a working example workflow.

Frequently Asked Questions

A strong README answers, in order, what the project does, why it matters, how to install it, and how to use it, with a runnable example near the top. These prompts also add badges, a table of contents, and the extras that fit the project type, such as an API reference, an env-var table, or a contribution guide.
Yes. Paste your package.json, action.yml, variables.tf, exported function signatures, or --help output into the input fields and Claude documents your real API instead of guessing. The more concrete detail you give it, the more accurate the install commands, examples, and tables will be.
Every prompt asks for a shields.io badge row appropriate to the project (version, downloads, build, license) and a table of contents. You just swap the placeholder package name or username so the badge and widget URLs point to your project.
They are ready to commit once you replace the bracketed placeholders with your real values. Claude returns valid GitHub-flavored Markdown with fenced code blocks and correct language tags, so you can paste it straight into README.md and preview it on GitHub.
Yes. GitHub natively renders fenced mermaid blocks, so the architecture and dependency diagrams display without any external image. Dynamic profile widgets like github-readme-stats render as images once you swap in your username, and Claude flags any that need a deployed instance or token.
Each prompt ends with a note on staying in sync, such as regenerating a command reference from --help, using terraform-docs for input/output tables, or generating schema docs. For fast-moving projects, re-run the relevant prompt with updated inputs whenever your API or config changes.

Prompts are the starting line. Tutorials are the finish.

A growing library of 300+ hands-on tutorials on ChatGPT, Claude, Midjourney, and 50+ AI tools. New tutorials added every week.

7-day free trial. Cancel anytime.