repo-meta
Skill0 starsChanged 8 months agoMITView on GitHub
Read the current working folder and suggest a GitHub repo name, description, topic tags, and an image-generation prompt for a logo. Use when the user wants to publish a project to GitHub, or invokes /repo-meta.
In this repository8 other files it gives its agents.
- releaseSkill
- fix-englishSkill
- mac-llm-fitSkill
- pushSkill
- research-apisSkill
- run-appSkill
- write-user-guideSkill
- write-user-storySkill
The file
We split this skill into the rules it gives an agent
9 rules. About 0 are common, 8 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈8
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 17
If signals are thin or conflicting, ASK one short clarifying question (e.g.
- Very specific: line 45
Avoid trademarked names (don't prefix with react-, next-, openai-, etc.
- Distinctive: line 46
Don't reuse the user's GitHub username as a prefix unless it's already their convention in the folder.
- Distinctive: line 51
No marketing fluff ("blazing fast", "next-gen").
- Distinctive: line 56
Use tags that already exist on GitHub (check https://github.com/topics/{tag} mentally — prefer common ones over invented ones).
- Distinctive: line 59
Write ONE paragraph, 2–4 sentences, that an image model can consume directly.
- Distinctive: line 60
Must specify: subject, style (flat / 3D / isometric / minimalist / etc.), color palette, and a solid background color (because the user …
- Distinctive: line 61
Keep the subject simple — a single icon or symbol, not a scene.
- Distinctive: line 63
Do NOT include text or letters in the logo (image models render text poorly).
≈ “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
Repo Meta
Inspect the current working directory and propose GitHub metadata the user can paste straight into the "Create repository" form, plus a ready-to-use logo prompt for an image model (e.g. Gemini Nano Banana, Midjourney, DALL·E).
Process
- Detect the project. Read only what's needed — in priority order, stop once you have enough:
README.md(title, tagline, first paragraph) — usually the best single source- The primary manifest for the stack:
package.json,pyproject.toml,Cargo.toml,go.mod, or.claude-plugin/ - Top-level folder names only if the above are thin
- If signals are thin or conflicting, ASK one short clarifying question (e.g. "Is this a CLI, library, or web app?") before generating suggestions. Otherwise skip straight to output.
- Produce the output below. Offer 3 repo name options so the user can pick one.
Output Format
**Detected:** {one-line summary — e.g. "Node.js CLI for converting Markdown to PDF"}
**Repo name options:**
1. `{kebab-case-name-1}` — {why it fits}
2. `{kebab-case-name-2}` — {why it fits}
3. `{kebab-case-name-3}` — {why it fits}
**Description** (80–100 chars, GitHub "About" field):
> {one-sentence description, no trailing period unless it reads better with one}
**Topics** (GitHub tags, lowercase, hyphenated, max 20):
`tag-1` `tag-2` `tag-3` ...
**Logo prompt** (for image AI, non-transparent background):
> {prompt paragraph — see Rules below}Rules
Repo name
- Lowercase, kebab-case, ASCII only, no leading/trailing hyphens.
- Short (ideally ≤ 25 chars). Memorable over descriptive.
- Avoid trademarked names (don't prefix with
react-,next-,openai-, etc. unless the project is genuinely an extension of that ecosystem). - Don't reuse the user's GitHub username as a prefix unless it's already their convention in the folder.
Description
- One sentence, 80–100 characters — aim for this range, don't go over.
- Lead with the noun (what it IS), then the verb (what it does). Example: "CLI that converts Markdown to styled PDFs."
- No marketing fluff ("blazing fast", "next-gen"). No emojis unless the user explicitly asks.
Topics
- Lowercase, hyphen-separated, 5–15 tags.
- Mix: language (
typescript), framework (nextjs), domain (markdown,pdf), and type (cli,library,plugin). - Use tags that already exist on GitHub (check https://github.com/topics/{tag} mentally — prefer common ones over invented ones).
Logo prompt
- Write ONE paragraph, 2–4 sentences, that an image model can consume directly.
- Must specify: subject, style (flat / 3D / isometric / minimalist / etc.), color palette, and a solid background color (because the user generates non-transparent logos).
- Keep the subject simple — a single icon or symbol, not a scene.
- End with format hint:
Square 1:1, centered, suitable for a GitHub repo avatar. - Do NOT include text or letters in the logo (image models render text poorly).
Example
**Detected:** Node.js CLI that batch-renames image files using EXIF data
**Repo name options:**
1. `exif-rename` — direct, describes the core action
2. `photo-tidy` — friendlier, works if you expand beyond EXIF later
3. `snap-sort` — short and memorable
**Description:**
> CLI that batch-renames photos based on EXIF timestamps and camera metadata
**Topics:**
`cli` `nodejs` `typescript` `exif` `photography` `image-processing` `batch-rename` `metadata`
**Logo prompt:**
> A minimalist flat-design icon of a camera shutter blending into a file-folder shape, rendered in soft teal and warm orange on a solid deep-navy background. Clean geometric lines, subtle inner shadow for depth, no text. Square 1:1, centered, suitable for a GitHub repo avatar.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.

