[codex] Add visual spec package intake - #4
Conversation
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 749c0ec305
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| @@ -1,4 +1,4 @@ | |||
| --- | |||
| --- | |||
There was a problem hiding this comment.
This command template now starts with a UTF-8 BOM before the opening ---. In contexts that parse command metadata by checking for the front-matter delimiter at the start of the file, /speckit.intake.visual-design will be the only command in this extension whose description block is not recognized after installation, unlike the other command templates that begin directly with ---. Remove the BOM so the front matter starts at byte 0.
Useful? React with 👍 / 👎.
| if not components or missing_preview_refs: | ||
| blocker_codes.append(BLOCKERS["FIGMA_NODE_COVERAGE_INCOMPLETE"]) | ||
| if missing_records or missing_visual_spec_refs: | ||
| blocker_codes.append(BLOCKERS["COMPONENT_STATE_COVERAGE_INCOMPLETE"]) |
There was a problem hiding this comment.
This readiness gate only blocks when authors explicitly add missing records, but the schema does not enforce that every value in required_dimensions appears in covered. A component can declare required states such as default and hover, cover only default, leave missing: [], and still pass preview validation, even though the preview contract requires every required state/variant/viewport to be covered or recorded missing. Please derive the expected dimension coverage from required_dimensions and emit VISUAL_PREVIEW_COMPONENT_STATE_COVERAGE_INCOMPLETE for uncovered combinations.
Useful? React with 👍 / 👎.
1 participant
Summary
This PR refactors the visual-design intake flow around a single target artifact family: the visual requirements/spec structured asset package.
What changed
/speckit.intake.visual-designinstead of keeping a separate command./speckit.intake.irand/speckit.intake.figma2htmlssotcommand surfaces.visual-spec-package/artifacts, schemas, contract, evidence packet template, and validator.previews/as reviewer helper evidence withcomponent-matrix-preview.html,component-coverage.yaml, andviewport-coverage.yaml.Why
The goal is high-fidelity visual/UI delivery from design-source evidence. The target deliverable should be a structured visual requirements/spec asset package, while preview HTML remains only a human review mirror for coverage completeness.
Impact
Downstream workflows now consume
visual-spec-package/as the structured target. Preview coverage can still support review, but it no longer acts as a target implementation artifact or source of truth.Validation
python -m pytest -q tests/test_extension_contract.py -k "visual_previews or visual_spec_package or manifest_loads or config_template"-> 30 passedpython -m pytest -q-> 71 passedgit diff --check --cached-> passed