agentleFS
samulee003samulee003/agentic-ppt-skills/.agents/skills/make-presentation/SKILL.md

make-presentation

Skill0 starsChanged 24 days agoMITView on GitHub

Turns a source folder, existing deck, or presentation idea into a complete presentation product by routing through source intake, product grilling, evidence, story, prototype, stage review, and final delivery. Use when starting a presentation from materials, rebuilding a deck, resuming a partial presentation project, or asking for an end-to-end PPT/PDF workflow.

The file

Show its rules and text

We split this skill into the rules it gives an agent

16 rules. About 0 are common, 5 are distinctive, and 11 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈5

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 11

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Read repo instructions first.

    line 8
  2. Very specific:

    Use the shipped helper at .agents/skills/make-presentation/scripts/status.mjs.

    line 20
  3. Very specific:

    If presentation-work/<deck-id>/STATUS.md exists, run node .agents/skills/make-presentation/scripts/status.mjs validate …

    line 21
  4. Very specific:

    Never reset an existing status file unless the user explicitly requests a reset.

    line 23
  5. Distinctive:

    Treat artifact existence as a clue, not proof that a gate passed.

    line 24
  6. Very specific:

    Run node .agents/skills/make-presentation/scripts/status.mjs next <work-root> and resume from the returned earliest unresolved gate.

    line 25
  7. Very specific:

    Follow references/status-schema.md for gate names, status values, and update commands.

    line 26
  8. Very specific:

    Read references/modes.md before routing.

    line 30
  9. Very specific:

    Read the current mode from STATUS.md before invoking Evidence or Finalization.

    line 36
  10. Very specific:

    Read references/gates.md for pass conditions and artifact owners.

    line 41
  11. Distinctive:

    Do not skip ahead because a PPTX, PDF, or other artifact already exists.

    line 52
  12. Very specific:

    Read references/gate-results.md, then translate a passed focused-skill result with

    line 53
  13. Very specific:

    Use node .agents/skills/make-presentation/scripts/status.mjs set when recording a reviewed artifact path, inputs, or decisions outside the …

    line 54
  14. Very specific:

    If product decisions remain unresolved, stop and grill in rounds per references/grilling-rounds.md: the whole frontier per round, each …

    line 55
  15. Distinctive:

    If upstream decisions change, invalidate downstream gates and resume from the earliest invalidated gate.

    line 56
  16. Distinctive:

    Read the named reference only when its trigger applies.

    line 60

≈ “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.

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

Make a presentation product

Read repo instructions first. Route the project from the earliest unresolved gate instead of recreating work.

Workspace

  • Derive or confirm deck-id, then use presentation-work/<deck-id>/.
  • Original presentation source folders are read-only by default.
  • Inspect existing artifacts in the source folder, work area, and slide folder before creating anything new.
  • Reuse equivalent artifacts when they already satisfy the current gate.
  • Writing artifacts: every gate that specifies writing an artifact from a template (e.g. from a named template path) means create that file in the work area using your runtime's file-write tool (the Write/Edit tool, or shell redirection if that is all you have). Start from the named template, fill it with real content for this deck, and write it to the path the gate specifies. There is no separate build script — file creation is your responsibility.

Status

  • Use the shipped helper at .agents/skills/make-presentation/scripts/status.mjs.
  • If presentation-work/<deck-id>/STATUS.md exists, run node .agents/skills/make-presentation/scripts/status.mjs validate presentation-work/<deck-id>.
  • Otherwise initialize it with node .agents/skills/make-presentation/scripts/status.mjs init <work-root> <deck-id> <source-root> <mode> <mode-reason>.
  • Never reset an existing status file unless the user explicitly requests a reset; only then append --force.
  • Treat artifact existence as a clue, not proof that a gate passed.
  • Run node .agents/skills/make-presentation/scripts/status.mjs next <work-root> and resume from the returned earliest unresolved gate.
  • Follow references/status-schema.md for gate names, status values, and update commands.

Mode

Read references/modes.md before routing.

  • Choose general for internal, creative, or low-risk decks without consequential external claims.
  • Choose fast-track for lightweight, agile decks (< 10 slides) when speed is prioritized.
  • Choose research-enhanced when the deck depends on consequential external claims, research, citations, datasets, policy, law, medicine, psychology, finance, safety, or statistics.
  • Re-evaluate mode after source intake. If later evidence triggers research safeguards, run node .agents/skills/make-presentation/scripts/status.mjs set-mode <work-root> research-enhanced "<reason>".
  • Read the current mode from STATUS.md before invoking Evidence or Finalization.
  • Under deadline pressure, reduce scope; never silently waive evidence, readability, or export integrity.

Route one gate at a time

Read references/gates.md for pass conditions and artifact owners. Invoke exactly one focused skill for the earliest unresolved gate:

  1. source-intake → presentation-source-intake
  2. product-grill → presentation-product-grill
  3. evidence → presentation-evidence
  4. story-architecture → presentation-story-architecture
  5. story-grill → presentation-story-grill
  6. deck-prototype → presentation-prototype
  7. stage-grill → presentation-stage-grill
  8. production-qa → presentation-deck-qa
  • Do not skip ahead because a PPTX, PDF, or other artifact already exists.
  • Read references/gate-results.md, then translate a passed focused-skill result with node .agents/skills/make-presentation/scripts/status.mjs result <work-root> <gate> passed "<reason>" --artifact <path> [--artifact <path>]... [--input <value>]... [--decision <value>]....
  • Use node .agents/skills/make-presentation/scripts/status.mjs set when recording a reviewed artifact path, inputs, or decisions outside the result handoff.
  • If product decisions remain unresolved, stop and grill in rounds per references/grilling-rounds.md: the whole frontier per round, each question with a recommended answer.
  • If upstream decisions change, invalidate downstream gates and resume from the earliest invalidated gate.

Cross-cutting practices

Read the named reference only when its trigger applies.

  • Grilling rounds (references/grilling-rounds.md): every grill gate asks the whole frontier of settled-prerequisite decisions per round, with "yes"-accepts recommendations, instead of one question per turn.
  • Sharpen and play back (references/sharpen-playback.md): at product-grill, play the goal back in one paragraph and proceed on recommended defaults for reversible choices instead of asking every question.
  • Unslop (references/unslop-copy.md): at story-architecture, deck-prototype, stage-grill, and production-qa, run node .agents/skills/make-presentation/scripts/unslop-check.mjs <path>... [--strict] on slide copy and speaker notes, then rewrite by hand.
  • Review buckets (references/review-buckets.md): at story-grill, stage-grill, and production-qa, sort every review finding into Act on, Consider, Noted, or Dismissed.
  • Arena (references/arena-prototype.md): at deck-prototype, build 2 to 3 competing directions for high-stakes or unsettled visual direction, pick a base, and graft the best parts. Skip in fast-track.
  • Encode lessons (references/encode-lessons.md): when the user corrects the same mistake twice, fix it in the template or a check first, a test next, and a doc last. Keep the rule table current.

Handoff

Report the chosen mode, selected engine (once resolved at prototype), current gate, reused artifacts, and next focused skill.

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.