agentsclimarketplace

Demo script architect

Skill ralvarezdev/ralvaskills/skills/personal/demo-script-architect

My personal, ever-growing collection of AI skills for OpenCode and Claude Code. Enforces strict clean architecture and professional standards.

Install
npx -y skills add ralvarezdev/ralvaskills --skill demo-script-architect

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 2 stars2 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Design and refine presenter-centric demo scripts — narrative flow over feature lists, visual-cue mapping, progressive capability reveals, query-to-explanation alignment, redundancy audits. Use when creating live demos, structured walkthroughs, or turning specs into engaging narratives.

SKILL.md

5.1 KB, as published. Nobody here has run it

Demo Script Architect

Design and refine presenter-centric demo scripts that guide audiences through system capabilities in a natural, engaging narrative flow. Works for any technical or product demo.

When to use

Creating live demonstration scripts; building structured walkthroughs of systems/features; designing progressive capability reveals; turning technical specs into engaging narratives; adding visual guidance for presenters.

1. Core principles

  • Narrative over features. Each section builds on previous learning; bridge narration explains the why. Story arc: observe → understand → apply → verify. Conversational, not documentation-style.
  • Element documentation. Always include element purpose when referenced: YI-100 (pump status indicator). Help the audience understand what they're looking at.
  • Visual guidance mapping. Map abstract references to actual visible elements. Data field "status" → point to the status icon/light. Never point to something the audience can't see on screen.
  • Query-to-explanation alignment. The explanation must accurately describe what's being queried — don't claim "context-learning" if tags are explicit. Explain the mechanism (parallelization, routing, lookup) and set up expectation for the response.
  • No redundancy. Don't repeat the same query/capability in consecutive demos. Build progressive revelation: basic → complex → combined → edge cases. Stack elements only when combining for a new insight.
  • Conversational tone. Use contractions ("it's", "we've"). Direct the audience ("Watch", "Notice", "Observe"). Short punchy sentences over long lists. Tell a story.
  • Commentary interprets, doesn't restate. Explains significance, not mechanics; sets up the next logical question.

2. Narrative section structure

Every demo follows the same six-part shape:

Demo Setup         → purpose, what to watch for
Visual Cue         → "[On screen: Point to X, watch Y]"
Action/Query       → explicit ask of the system
Processing         → 2-3 audience-friendly sentences on what's happening
Expected Responses → multiple outcomes based on system state
Commentary         → interpret results, relate to broader picture

3. Demo progression template

#FocusNew capabilityApproach
1BasicsSingle element accessExplicit reference
2IntegrationMultiple related elementsGrouped query
3AbstractionNatural language, no explicit refsSystem infers intent
4DynamicsLive data, change over timeContinuous monitoring
5AccessibilityDifferent access methodsVoice/touch/alternative interface
6AutomationSystem behavior without commandObserve automatic logic
7DiagnosticsTroubleshooting scenarioBundle related elements
8KnowledgeSearch/lookup capabilityFind answers in resources
9ControlIssue commandsWrite operations + verification

4. Anti-patterns

Anti-patternWrongRight
Explanation/query mismatchQuery mentions YI-100, SI-100, PIT-100; explanation says "we didn't mention specific tags this time""Three different data points from two systems in parallel"
Pointing to invisible"[Point to field PIT-100 on screen]" when not labeled"[Point to the pressure gauge display]"
Robotic narration"System processes query, retrieves values from database, returns formatted response.""Watch the system grab the values and hand them back as one answer."
Redundant demosDemo 1 status query → Demo 2 same status queryDemo 1 status query → Demo 2 combine status with related data
Weak bridge"Now let's do a different demo.""Now that we know the status, let's check the pressure to see what it's delivering."

5. Review checklist

  • Each demo has a Setup explaining purpose
  • Actions/Queries are explicit and clear
  • Processing accurately describes what's happening
  • Visual cues point to visible screen elements
  • Element names include purpose: NAME (what it does)
  • Commentary interprets results
  • Bridge narration connects each demo to the next with a why
  • No redundant queries (same action shown twice for same purpose)
  • Multiple expected responses cover realistic system states
  • Tone is conversational throughout (no robotic jargon)

6. Output deliverables

  1. Structured script with the six-part narrative sections (§2).
  2. Visual element map — reference → visible screen element.
  3. Demo flow diagram showing connections and progression.
  4. Redundancy audit — what was removed/consolidated and why.
  5. Presenter notes — what to watch for, where to point, what to emphasize.

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.