ft-readme-style
Skill1 starsChanged 3 months agoMITView on GitHub
Write or simplify project READMEs into a fast-scanning project front page: centered logo, compact badges, a clear explanation, short examples, and links to focused supporting documents. Use when creating, rewriting, or reviewing a README.
In this repository22 other files it gives its agents.
- skills/vercel-composition-patterns/AGENTS.mdAGENTS.md
- skills/vercel-react-best-practices/AGENTS.mdAGENTS.md
- agent-browserSkill
- design-taste-frontendSkill
- fal-image-genSkill
- find-skillsSkill
- ft-netbird-shareSkill
- ft-visual-styleSkill
- gitadelSkill
- impeccableSkill
- react-bits-proSkill
- react-doctorSkill
- researchSkill
- rust-best-practicesSkill
- scorchSkill
- shell-secretsSkill
- supermemory-auditSkill
- svelte-code-writerSkill
- svelte-core-bestpracticesSkill
- unslopSkill
- vercel-composition-patternsSkill
- vercel-react-best-practicesSkill
The file
Show its rules and text
We split this skill into the rules it gives an agent
14 rules. About 0 are common, 12 are distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈12
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 12
Start with a centered project logo or mark, then place a compact row of relevant badges directly below it.
- Distinctive: line 26
Use only badges that answer a decision-relevant question: release/version, CI health, license, package registry, or documentation.
- Distinctive: line 36
Prefer clear headings, short paragraphs, and concrete verbs.
- Distinctive: line 37
Remove background narrative, implementation archaeology, exhaustive option tables, and features readers cannot act on.
- Distinctive: line 38
Avoid marketing claims that do not explain behavior.
- Distinctive: line 39
Do not make users scroll through duplicated information.
- Distinctive: line 40
Write for a first-time visitor, not the maintainers who already know the project.
- Very specific: line 44
Put full installation, prerequisites, platform matrices, environment setup, and troubleshooting in INSTALL.md or an existing installation …
- Very specific: line 45
Put contributor setup, development commands, architecture guidance, release process, and coding rules in CONTRIBUTING.md or the …
- Distinctive: line 46
Put API/reference material, security policy, changelog, and design rationale in their own established documents.
- Distinctive: line 47
Create a new supporting document only when the project lacks a suitable home and the detail is necessary.
- Distinctive: line 58
Keep the number of compared items small.
- Distinctive: line 61
Never bury the reader in benchmark harness boilerplate, raw logs, or a wall of qualifiers.
- Distinctive: line 65
Before finishing, verify that the README answers these questions without unnecessary reading:
≈ “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
FT README Style
A README is a project's front door, not its complete manual. A reader should understand the project, whether it matters to them, and what to try next in under a minute.
Required opening
Start with a centered project logo or mark, then place a compact row of relevant badges directly below it.
<p align="center">
<img src="assets/logo.svg" alt="Project name" width="160" />
</p>
<p align="center">
<a href="..."><img src="..." alt="CI status" /></a>
<a href="..."><img src="..." alt="License" /></a>
<a href="..."><img src="..." alt="Version" /></a>
</p>Use only badges that answer a decision-relevant question: release/version, CI health, license, package registry, or documentation. Do not add vanity counters or a badge wall.
Follow with:
- A one-sentence plain-language definition of the project.
- A short paragraph explaining the problem, distinctive approach, and intended user.
- At most three concise bullets for core capabilities when they improve scanning.
Keep the README small
- Prefer clear headings, short paragraphs, and concrete verbs.
- Remove background narrative, implementation archaeology, exhaustive option tables, and features readers cannot act on.
- Avoid marketing claims that do not explain behavior.
- Do not make users scroll through duplicated information.
- Write for a first-time visitor, not the maintainers who already know the project.
Move detail to focused documents
- Put full installation, prerequisites, platform matrices, environment setup, and troubleshooting in
INSTALL.mdor an existing installation guide. Link to it from the README; do not duplicate it. - Put contributor setup, development commands, architecture guidance, release process, and coding rules in
CONTRIBUTING.mdor the repository's established contributor guide. - Put API/reference material, security policy, changelog, and design rationale in their own established documents.
- Create a new supporting document only when the project lacks a suitable home and the detail is necessary.
Examples
Include one to three examples only when they prove the primary workflow. Each example must be runnable or directly understandable without surrounding prose. Lead with the smallest useful invocation or code sample; link out for every advanced case.
Charts, comparisons, and benchmarks
Include a chart, table, or benchmark only when it answers one clear question. Make the conclusion legible before the methodology:
- Name the measurement and units in the title or column heading.
- Keep the number of compared items small.
- Highlight the meaningful delta, not decorative styling.
- State the tested version, environment, and command in one compact note or a link to the methodology.
- Never bury the reader in benchmark harness boilerplate, raw logs, or a wall of qualifiers.
Final review
Before finishing, verify that the README answers these questions without unnecessary reading:
- What is this?
- Who is it for and why would they use it?
- What is the fastest credible example?
- Where are installation and contributor details?
- What evidence supports any performance or compatibility claim?
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.

