agentleFS
imMamdouhaboammarimMamdouhaboammar/linkedin-animated-infographics/skills/share-demo/SKILL.md

share-demo

Skill14 starsChanged 3 months agoMITView on GitHub

Package a verified infographic demo for the public gallery and, after explicit user opt-in, prepare a community pull request without merging it.

The file

We split this skill into the rules it gives an agent

6 rules. About 0 are common, 4 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    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
  1. Distinctive:

    Run the focused parent workflow for publishing a finished infographic to the repository community gallery.

    line 12
  2. Distinctive:

    Do not invoke this workflow merely because a GIF was generated.

    line 20
  3. Very specific:

    Read build/verification-report.json.

    line 56
  4. Distinctive:

    Treat the source prompt as private by default.

    line 60
  5. Distinctive:

    Write the approved metadata input used by the deterministic preflight.

    line 61
  6. Very specific:

    Run the gallery validator against the contribution checkout before GitHub publication.

    line 79

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

/linkedin-animated-infographics:share-demo

Purpose

Run the focused parent workflow for publishing a finished infographic to the repository community gallery. This is a public data-export boundary, not part of normal delivery. Read helper/GUIDE.md before execution and keep all GitHub writes behind explicit opt-in.

The public demo contract is exactly demo.gif + index.html + demo.json. demos/catalog.json is generated separately and is never a fourth file inside a demo directory.

Use when

Use this focused workflow only when a user explicitly agrees to share a finished demo with the community, or directly asks to publish a previously verified result. The normal new-post parent workflow may offer this after delivery only when final verification PASS has already been recorded.

Do not invoke this workflow merely because a GIF was generated. Declining the offer, giving no consent, or asking only for local output causes no GitHub write.

Inputs

Required inputs:

  • explicit opt-in to public community sharing
  • build/post.html or another explicitly selected final verified HTML artifact
  • build/post.gif or another explicitly selected final verified GIF artifact
  • build/verification-report.json with final verdict PASS
  • GitHub username and matching public author URL
  • demo title and short description
  • language, story type, and tags
  • explicit rights confirmation covering the submitted HTML, GIF, and redistributable assets
  • chosen public license

Optional inputs:

  • source prompt, only when the user separately consents to publishing it
  • source repository URL
  • short public notes

Outputs

The parent workflow produces:

  • build/community-demo.json containing the approved public metadata
  • build/community-demo-package containing the staged three-file package
  • on successful GitHub publication, a pull request URL against imMamdouhaboammar/linkedin-animated-infographics:main
  • on unavailable GitHub write access, a precise HOLD plus the validated local contribution package and manual commands when the host can prepare them safely

A PR URL is required before reporting GitHub publication as successful. Creating files locally is not equivalent to opening a pull request.

Procedure

  1. Confirm explicit opt-in. Ask one direct question if consent is not already explicit. Do not infer consent from enthusiasm, prior GitHub use, or the fact that the user created the work.
  2. Read build/verification-report.json. Community publication requires verification PASS. Any other verdict returns a HOLD.
  3. Resolve the exact final HTML and GIF. Do not publish drafts, stills, evidence files, build reports, attachments, or unrelated files.
  4. Collect publication metadata: title, author, author URL, description, language, story type, tags, date, plugin version, and license.
  5. Obtain explicit rights confirmation. Do not infer ownership or redistribution rights.
  6. Treat the source prompt as private by default. Include it only after a separate explicit yes to publishing the prompt itself.
  7. Write the approved metadata input used by the deterministic preflight. Keep control fields such as prompt-consent flags outside the final public schema.
  8. Prepare the contribution package with scripts/demo_submit.py prepare, for example:
python3 scripts/demo_submit.py prepare \
  --build-dir build \
  --stage-root build/community-submission \
  --author <github-user> \
  --slug <slug> \
  --metadata build/community-demo.json
  1. Validate the staged package with scripts/demo_submit.py check:
python3 scripts/demo_submit.py check build/community-submission/community/<github-user>/<slug>
  1. Run the gallery validator against the contribution checkout before GitHub publication. The required gate is scripts/demo_gallery.py check; regenerate demos/catalog.json deterministically with scripts/demo_gallery.py build only after the three-file package is placed in the contribution checkout.
  2. Confirm the contribution changes only demos/community/<github-user>/<slug>/{demo.gif,index.html,demo.json} plus the root generated demos/catalog.json when its content changes.
  3. Delegate the validated contribution to community-publisher. That worker owns fork, branch, commit, push, and pull-request mechanics. It returns control to this parent workflow and never merges.
  4. Return the PR URL, demo ID/path, validation summary, and the explicit reminder that maintainer manual review and merge are still required.

HOLD conditions

Return HOLD instead of guessing or partially publishing when any of these is unresolved:

  • explicit opt-in is missing
  • final verification PASS is missing
  • final GIF or HTML is missing, empty, or ambiguous
  • rights confirmation is missing or false
  • public metadata is incomplete
  • secret, local-path, signed-URL, or unsafe remote-script findings appear in export preflight
  • the destination demo already exists
  • GitHub username cannot be resolved safely
  • GitHub authentication cannot create/reuse a fork, push a fresh branch, or open a pull request
  • the target upstream repository cannot be verified exactly
  • deterministic catalog validation fails

A HOLD must state the blocking requirement. Never weaken export checks, publish extra files, or claim success without a returned PR URL.

Related components

  • routing authority: helper/GUIDE.md
  • route registry: helper/router.json
  • demo schema: schemas/demo.schema.json
  • package preflight: scripts/demo_submit.py
  • gallery validator: scripts/demo_gallery.py
  • gallery index: demos/catalog.json
  • publisher worker: agents/community-publisher.md
  • final creation parent workflow: skills/new-post/SKILL.md
  • artifact registry: helper/artifacts.json

Research gates

bounded-verification remains the relevant research gate because a demo cannot cross this export boundary until the independent final verdict is PASS. Existing evidence and creative gates govern the artifact before publication; this workflow does not create new claims, redesign the post, or reinterpret evidence.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.