cli-forge-prez
Skill5 starsChanged 8 months agoMITView on GitHub
Generate technical presentation decks in Marp Markdown from code, architecture, or documentation. Produces speaker-ready slides with notes, dark theme, and CI templates for PDF/HTML export. Use when the user mentions 'prez', 'presentation', 'slides', 'deck', 'talk', 'lightning talk', 'pitch', 'conference talk', 'demo day', 'speaker notes', 'technical presentation', 'Marp', 'slide deck', or wants to present a project, architecture, or design to an audience. Also triggers on 'keynote', 'meetup talk', 'sprint review prez', 'make slides', 'present this'.
1 other project uses a version of this, 1 of them edited.
Adopted by
Most recently taken up by
- murillo-consulting/cli-skillsedited, first seen Oct 20, 2026
Other versions1 project uses a version of this.
Word for word · 0
None found.
Edited · 1
- murillo-consulting/cli-skills★ 099% same
Wrote their own version · 0
None found.
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository40 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- .cursor/rules/cli-audit-code.mdcCursor rule
- .cursor/rules/cli-audit-doc.mdcCursor rule
- .cursor/rules/cli-forge-arch.mdcCursor rule
- .cursor/rules/cli-forge-schema.mdcCursor rule
- cli-audit-codeSkill
- cli-audit-dataSkill
- cli-audit-docSkill
- cli-audit-driftSkill
- cli-audit-hanoiSkill
- cli-audit-reviewSkill
- cli-audit-shellSkill
- cli-audit-syncSkill
- cli-audit-tangleSkill
- cli-audit-testSkill
- cli-audit-wizardSkill
- cli-audit-xraySkill
- cli-cycleSkill
- cli-forge-archSkill
- cli-forge-chefSkill
- cli-forge-choice-uxSkill
- cli-forge-dataSkill
- cli-forge-demoSkill
- cli-forge-docSkill
- cli-forge-githubSkill
- cli-forge-hldSkill
- cli-forge-infraSkill
- cli-forge-lldSkill
- cli-forge-oci-rootlessSkill
- cli-forge-perfSkill
- cli-forge-pipelineSkill
- cli-forge-plumeSkill
- cli-forge-profileSkill
- cli-forge-quorumSkill
- cli-forge-readmeSkill
- cli-forge-resilienceSkill
- cli-forge-schemaSkill
- cli-forge-treeSkill
- cli-git-conventionalSkill
- cli-watermarkSkill
The file
We split this skill into the rules it gives an agent
11 rules. About 0 are common, 3 are distinctive, and 8 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈3
Distinctive · estimate
This author's own choices. Worth a look.
8
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 57
Ask the user or take it from arguments.
- Very specific: line 59
Read references/tiers.md for the three tiers:
- Very specific: line 69
Read the project. Build a mental model of: what it does, who it's for, what problem it solves, what makes it different.
- Very specific: line 73
Read references/slide-structure.md for frameworks by tier.
- Very specific: line 80
Read references/marp-reference.md for syntax, themes, and advanced techniques.
- Distinctive: line 83
Every slide passes the 5-second test — the audience grasps the message in 5 seconds
- Very specific: line 84
Every slide has speaker notes in <!-- ...
- Very specific: line 93
Write the .md file to docs/slides/{tier}-{name}.md (or talks/ if the project convention differs).
- Very specific: line 95
If the project has CI (GitHub Actions or GitLab CI), suggest the CI template from references/ci-templates.md for PDF generation and Pages …
- Very specific: line 111
Read references/anti-patterns.md for the 11 named anti-patterns to detect and avoid during generation (Feature List, Wall of Code, …
- Distinctive: line 126
The slides support the speaker, they don't replace them: If the deck is readable without a speaker, it's a document — not a presentation.
≈ “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
Optimization: This skill uses on-demand loading. Heavy content lives in
references/and is loaded only when needed.
Language rule: Skill instructions are English. Detect the project's primary language (from README, comments, docs, commit messages). Generate the presentation in that language. If bilingual, ask the user.
Diacritics rule: When the chosen output language uses diacritics (French, Spanish, Portuguese, etc.), render them correctly — including on capitals (É, À, Ç, Ô…). Never strip accents to ASCII: write "sécurité" not "securite", "Rôle" not "Role". ASCII-flattened French is a defect, not a style.
Writing style: Favor clarity — short sentences, one idea per sentence, plain words. A metaphor only when it clarifies. Keep one consistent register (impersonal/infinitive for operational docs). Separate narrative prose, meant to make the reader understand, from telegraphic content (tables, commands, checklists) which stays terse for action.
No fabricated output: Never invent command output, console captures, API responses, metrics, or logs. When real output is needed, emit a clearly-marked placeholder (🚧 "replace with a real capture"), never plausible-but-fake text presented as real.
Presentation Forge — Marp Deck Generator
Generate speaker-ready technical presentation decks from any project artifact — code, architecture docs, ADRs, audit reports, or a plain description.
Core Principle — Two biological mechanisms
| Mechanism | What it does in nature | What it does in this skill |
| Peacock display | Minimum viable signal, maximum recognition distance. Each ocelle is a high-contrast circle — one shape, instantly recognized by the peahen from across the clearing | One message per slide, instantly understood from the back of the room. If you must squint (mentally or physically), the signal failed |
| Pheromone trail | Ants lay a chemical trail between food and nest; each waypoint reinforces the path. Remove a waypoint and the colony detours; add a false one and ants get lost | The narrative arc is a trail the audience follows slide by slide. Each slide is a waypoint. Remove it: if the trail breaks, it was load-bearing. If nobody notices, it was noise — delete it |
Input
$ARGUMENTS is one of:
- File or directory: the skill reads the project and builds a presentation about it
- Topic description: "pitch for grob — LLM proxy with audit trail"
- Empty: reads the current project root and asks which tier
Workflow
Step 0 — Determine tier
Ask the user or take it from arguments. Do not auto-detect — presentation context is too varied to infer.
Read references/tiers.md for the three tiers:
| Tier | Context | Slides | Duration |
| pitch | Elevator, README hero, hallway | 3-4 | ~2 min |
| lightning | Meetup, demo day, sprint review | 6-8 | ~5 min |
| talk | Conference, workshop, client prez | 10-12 | ~20 min |
Step 1 — Reconnaissance
Read the project. Build a mental model of: what it does, who it's for, what problem it solves, what makes it different. Use the shared reconnaissance brief — read ../shared/recon.md (the same model readme and demo use, so all three tell one story).
Step 2 — Choose narrative framework
Read references/slide-structure.md for frameworks by tier. Default:
- pitch: Hook → Solution → Proof → CTA
- lightning: SCR (Situation-Complication-Resolution)
- talk: 5-Box (What → Why → How → What If → What Next)
Step 3 — Generate Marp Markdown
Read references/marp-reference.md for syntax, themes, and advanced techniques.
Mandatory rules:
- Every slide passes the 5-second test — the audience grasps the message in 5 seconds
- Every slide has speaker notes in
<!-- ... -->comments — not optional - Headings are assertions, not labels (Alley model):
# Pods restart 3x faster with preemptionnot# Preemption - Max 6 words on visual slides, max 5 bullet points on content slides
- Code blocks: max 8 lines, highlight the 2-3 lines that matter
- Dark theme by default (better for conference projectors in bright rooms)
- Mermaid diagrams via
cli-forge-schemahandoff when architecture visuals are needed
Step 4 — Write output files
Write the .md file to docs/slides/{tier}-{name}.md (or talks/ if the project convention differs). Standard path, no AI markers.
If the project has CI (GitHub Actions or GitLab CI), suggest the CI template from references/ci-templates.md for PDF generation and Pages deployment.
Step 5 — Quality checklist
Before delivering, verify:
- Every slide passes the 5-second rule (peacock test)
- Removing any slide either breaks the narrative trail or proves it was noise
- Speaker notes present on every slide
- No slide has > 8 lines of code or > 5 bullet points
- Opening hook and closing slide form a callback pair
- Dark theme renders well (high contrast, no color-only information)
- Total slide count matches tier expectation
Anti-Patterns
Read references/anti-patterns.md for the 11 named anti-patterns to detect and avoid during generation (Feature List, Wall of Code, Architecture Astronaut, Demo Gods, Slideument, Curse of Knowledge, Apology Slide, Bibliography, Outline Slide, Font Roulette, Premature Abstraction).
Dynamic Handoffs
| Condition detected | Recommend | Why |
| Architecture diagram needed | /cli-forge-schema | Generate Mermaid for embedding in slides |
| Presenting audit results | /cli-audit-code or /cli-audit-doc | Run the audit first, then present findings |
| Presenting project structure | /cli-forge-tree | Generate annotated tree for slides |
| Presenting an HLD/LLD | /cli-forge-hld or /cli-forge-lld | Generate design doc, then distill into slides |
Rule: Recommend, don't auto-execute.
Rules for the Generator
- The slides support the speaker, they don't replace them. If the deck is readable without a speaker, it's a document — not a presentation.
- Concrete before abstract. Show the specific case, then generalize. Never the reverse.
- One idea per slide, one idea per talk (lightning). A lightning talk with two ideas is two bad talks.
- Speaker notes are the real content. The slides are visual anchors. The notes are what the speaker actually says.
- Dark theme by default. Conference projectors in bright rooms render dark backgrounds with light text better than the inverse.
- Gotchas — read
../gotchas.mdbefore producing output to avoid known mistakes.
Integration with other cli-* skills
| Skill | Relationship |
cli-forge-schema | Generates Mermaid diagrams embeddable in Marp slides |
cli-forge-readme | Both generate project descriptions; prez is the oral version, readme is the written one |
cli-forge-doc | Generates docs that can be distilled into slides |
cli-forge-hld / cli-forge-lld | Architecture docs → presentation material |
cli-audit-code / cli-audit-doc | Audit reports can become presentation content |
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.

