Rich Visualization Sketches

The five briefs in ../design/cycles/c38.qcontrol-visibility-storytelling/briefs/ are strong design prompts, but they ask bob-wire to express ideas that sit between wireframe and visualization. Today the project can represent those pages as generic panels and placeholders, but it does not yet have a reusable sketch language for:

• system shape
• evidence and provenance
• live vs accumulated state
• grouped entities and boundaries
• before / after comparison
• layered storytelling from raw signals to narrative

This cycle gives bob-wire that expression layer by treating qmap, qflow, and qpixel as reference grammars, then translating them into simple, static, design-review-friendly sketches.

Expand bob-wire beyond cards, tables, and form shells so it can sketch richer diagrammatic interfaces: relationship maps, data-flow views, grouped topology, truth / boundary overlays, provenance inspectors, abstraction ladders, and diff canvases.

1. Add a reusable visualization sketch skill for bob-wire agents
2. Teach wireframe and mockup builders when and how to use it
3. Create a bob-wire proving-ground brief set based on the five design c38 briefs
4. Define a shared visual grammar for relationships, data flow, grouping, authority, confidence, boundary, and diff
5. Establish a repeatable run model so the same briefs can be rerun across wf1, wf2, wf3 as the skill evolves

Out of scope:

• production implementation inside qmap, qflow, or qpixel
• final animation or interaction design for the visualization engines
• completing every proving-ground wireframe in this setup cycle

1. .claude/skills/visualization-sketches.md
2. Builder-agent guidance updates for wireframes and mockups
3. cycles/c8.rich-visualization-sketches/briefs/README.md
4. Five adapted proving-ground briefs in cycles/c8.rich-visualization-sketches/briefs/
5. cycles/c8.rich-visualization-sketches/RUNS.md — run protocol for repeated skill proofs

Bob-wire now has an explicit path for sketching rich visualization concepts instead of collapsing them into generic placeholders, plus a run model for rerunning the same briefs across evolving skill iterations.

Source cycle:

• ../design/cycles/c38.qcontrol-visibility-storytelling/CYCLE.md

Imported and adapted as wireframe-ready bob-wire briefs:

1. telemetry-boundary-layer.md
2. provenance-inspector.md
3. paired-live-session-surfaces.md
4. session-abstraction-ladder.md
5. session-diff-view.md

These briefs are the first test set for the new capability. If bob-wire can sketch them clearly, the expression layer is working.

• Cycle created
• Visualization sketch skill added
• Wireframe builder updated to use the new skill
• Mockup builder updated to preserve the same sketch grammar
• Design c38 proving-ground briefs adapted into bob-wire cycle briefs
• Versioned run model defined for repeated proofs against the same brief set
• wf1 completed for the full proving-ground brief set

• feature version answers: which brief / concept is being explored
• iteration answers: which proof run of the skill generated it

For c8, the proving-ground features stay stable at *-v1 while the skill is rerun across:

• wf1 — first proof pass
• wf2 — second proof pass after skill refinements
• wf3 — third proof pass, if needed

That lets the team compare the exact same brief across different generations of the visualization skill without conflating "brief changed" and "skill improved."