Finite VDC Engineer Challenge #1 — Make the Wall Editable

[jonathanmcmichael]

The challenge

Turn Rustit's first semantic wall from a rendered proof into the smallest credible authoring loop:

inspect semantic wall
        ↓
make a validated edit
        ↓
preserve stable identity
        ↓
regenerate neutral geometry
        ↓
see and identify the result

This is a coordination challenge, not permission to combine everything into one pull request. Claim one work packet, use whatever responsible AI coding workflow helps you, and return evidence through the normal review process.

Work packets

Author the semantic wall

• Add validated wall height editing #5 — validated wall-height editing
• Add validated wall baseline editing with stable identity #11 — validated baseline editing with stable identity
• Add keyboard-driven wall height editing to the desktop shell #23 — keyboard-driven height editing and regeneration in the desktop shell

Navigate and identify

• Add mouse-wheel zoom to the semantic wall renderer #8 — mouse-wheel zoom
• Add bounded orbit controls to the semantic wall view #15 — bounded orbit controls
• Map neutral mesh ranges back to stable element IDs #13 — map neutral mesh ranges to stable element IDs

Strengthen geometric truth

• Add a centered rectangular opening to Wall Lab #14 — centered rectangular-opening Wall Lab fixture

Challenge exit condition

Challenge #1 is complete when a fresh contributor can:

1. clone Rustit and run cargo xtask verify;
2. open the native application;
3. navigate around the wall;
4. make at least one validated semantic edit;
5. see regenerated geometry without changing the wall UUID; and
6. resolve rendered wall geometry back to that stable UUID.

Each behavior must be backed by a focused test or truth fixture. Interactive behavior also needs a reproducible live check.

How to participate

1. Read VISION.md, AGENTS.md, and the work packet you want.
2. Comment on that packet with the smallest outcome you plan to deliver.
3. Keep the pull request bounded to one packet.
4. Run cargo xtask verify.
5. Complete the pull-request evidence and AI-assistance disclosure.
6. Link discoveries as new packets rather than expanding scope.

Domain experts can contribute without writing Rust: review Wall Lab dimensions, interaction expectations, terminology, failure cases, and whether the result resembles a useful authoring workflow.

Scoreboard

Progress is measured by merged packets, strengthened fixtures, regressions caught, and returning contributors—not prompts, tokens, or generated lines.

No customer data, proprietary model files, credentials, or licensed classification catalogs may be used.