readme-creator
Skill1 starsChanged 6 months agoMITView on GitHub
Creates high-quality, standardized README.md files for software projects suitable for open-source repositories. Use when asked to create or update a README.
In this repository7 other files it gives its agents.
- braid-promptingSkill
- git-best-practicesSkill
- lucide-iconsSkill
- memgooseSkill
- milkdownSkill
- raison-sdkSkill
- reka-uiSkill
The file
We split this skill into the rules it gives an agent
11 rules. About 0 are common, 10 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈10
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 8
Create clear, professional README.md files grounded in the actual project.
- Distinctive: line 34
No invented commands, flags, or capabilities
- Distinctive: line 35
No filler text ("Feel free to...", "This project aims to...")
- Distinctive: line 36
No generic disclaimers or boilerplate paragraphs
- Distinctive: line 38
No ASCII diagrams — use Mermaid instead
- Distinctive: line 42
Use only information from:
- Distinctive: line 47
If information is missing, omit the section instead of guessing.
- Distinctive: line 56
Run the Validation Checklist before returning output.
- Distinctive: line 65
No marketing language ("powerful", "seamless", "blazing fast")
- Very specific: line 73
When a visual diagram adds value (architecture, flow, sequence, git graph, etc.), use a fenced Mermaid block ( `mermaid ).
- Distinctive: line 77
Before returning the README, verify:
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill 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
README Creator
Create clear, professional README.md files grounded in the actual project.
Required Sections (always include)
- Title + Description — Project name as
# Title, followed by 1–3 sentences: what it does and its purpose. No badges unless provided. - Features — Bulleted list of key capabilities. Be specific; don't invent features.
- Installation — Exact commands to install the project and its dependencies.
- Usage — Minimal working example(s) with code blocks. Show the most common use case first.
- Contributing — Brief guide: fork → branch → PR. One short paragraph or 3–4 bullets.
- License — State the license. Default to MIT if unspecified:
Distributed under the MIT License.
Conditional Sections (include only when justified)
- Prerequisites — Only if the project requires specific runtimes, accounts, environment variables, or non-obvious tools. Skip if setup is standard (e.g., just
npm install).
Optional Sections (include when they add clear value)
These sections are not limited to the items listed below. Add any section that would genuinely benefit readers of this specific project.
- Getting Started / Quickstart — When a more guided walkthrough helps beyond Installation + Usage.
- Documentation / Docs — When there's an external docs site or extensive API reference to link to.
- Learn More — When there are tutorials, blog posts, videos, or community resources worth linking to.
Constraints
- No "Project Structure" section
- No invented commands, flags, or capabilities
- No filler text ("Feel free to...", "This project aims to...")
- No generic disclaimers or boilerplate paragraphs
- Omit any section that would be empty or redundant
- No ASCII diagrams — use Mermaid instead
Input Sources
Use only information from:
- User description
- Provided files or repository contents
package.json, README drafts, or source code
If information is missing, omit the section instead of guessing.
Procedure
- Analyze the provided project information (description, source code,
package.json, etc.). - Identify the project's purpose, core features, and intended users.
- Extract real installation steps and usage examples from the project — do not invent them.
- Include all Required Sections.
- Include Conditional and Optional Sections only when justified.
- Run the Validation Checklist before returning output.
- Return only the final README.md content — no preamble, no explanation.
Style Guidelines
- Professional — clear, authoritative, not casual
- Direct — lead with what matters; no slow build-up
- Concise — one sentence where one sentence suffices
- Technical and precise — use correct terminology for the domain
- No marketing language ("powerful", "seamless", "blazing fast")
Output Format
- Standard Markdown, GitHub-flavored
- Fenced code blocks with language tags (
```bash,```js, etc.) - Flat heading hierarchy:
##for sections,###only when a section genuinely needs subsections - Blank line between sections for readability
- When a visual diagram adds value (architecture, flow, sequence, git graph, etc.), use a fenced Mermaid block (
```mermaid). Never use ASCII art for diagrams.
Validation Checklist (internal)
Before returning the README, verify:
- All Required Sections are present
- No forbidden sections included (e.g., Project Structure)
- No invented features, commands, or capabilities
- Conditional sections are included only when justified
- Markdown formatting is correct and consistent
- Content is concise and non-redundant
- No ASCII diagrams present (Mermaid used where diagrams appear)
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.

