Skip to content

chore: final-polish — remove video footprint + polish context docs#16

Merged
richsak merged 3 commits intodevfrom
chore/final-polish
Apr 27, 2026
Merged

chore: final-polish — remove video footprint + polish context docs#16
richsak merged 3 commits intodevfrom
chore/final-polish

Conversation

@richsak
Copy link
Copy Markdown
Owner

@richsak richsak commented Apr 27, 2026

Summary

Final polish pass before the dev → main rollup (PR #15).

  • Remove submission-tooling that doesn't belong in the public repo: skills/webster-video/, video/, context/webster-video/, the video composition prompt, the onboarding case-study, and the webster-onboarding empire fixture (~7000 line deletion).
  • Polish ARCHITECTURE / FEATURES / DOMAIN-MODEL / VISION / QUALITY-GATES to the post-eeda2bc shipped state — 9 production Managed Agents mirrored 1:1 by 9 simulation specs. Layer 6 (video) reframed as external submission tooling, not a blocked product layer. visual-design-critic correctly framed as a W4 genealogy spawn (its true provenance), not a permanent L2 base agent. Drop deferred / hackathon-crunch framing from the canon docs.
  • Strip unverifiable projections from README (cost-per-month, cost-per-run, agency-pricing comparisons). Standardize test-count phrasing.
  • Tidy dangling refs in skills/webster-onboarding and qa-bank.

Test plan

🤖 Generated with Claude Code

Remove submission-tooling that does not belong in the public repo:
skills/webster-video/, video/, context/webster-video/, the video
composition prompt, the onboarding case-study, and the
webster-onboarding empire fixture. The HyperFrames render pipeline
that produced the timelapse was always tooling-side, not product.

Polish ARCHITECTURE / FEATURES / DOMAIN-MODEL / VISION / QUALITY-GATES
to the post-eeda2bc shipped state: 9 production Managed Agents
mirrored 1:1 by 9 simulation specs. Reframe Layer 6 (video) as
external submission tooling, not a blocked product layer. Move
visual-design-critic to its true provenance — a W4 genealogy spawn,
not a permanent L2 base agent. Drop deferred / hackathon-crunch
language from the canon docs; that posture is over.

Strip unverifiable projections from README (cost-per-month,
cost-per-run, agency-pricing comparisons). Standardize the
test-count phrasing to "29 test files green via bun run validate".

Tidy dangling references in skills/webster-onboarding and qa-bank.

Validate is green: 176 tests pass, 0 lint warnings, 0 type errors,
0 markdown errors, 19 JSON specs valid, 7 findings files valid.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
@coderabbitai
Copy link
Copy Markdown

coderabbitai Bot commented Apr 27, 2026

Important

Review skipped

Auto reviews are disabled on base/target branches other than the default branch.

🗂️ Base branches to auto review (2)
  • main
  • review/empty-base

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

Run ID: be869a6f-2cea-48c1-a883-4538a3f26137

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

  • 🔍 Trigger review
✨ Finishing Touches
🧪 Generate unit tests (beta)
  • Create PR with unit tests
  • Commit unit tests in branch chore/final-polish

Comment @coderabbitai help to get the list of available commands and usage tips.

richsak and others added 2 commits April 26, 2026 18:01
Forge dispatch watcher with no live consumer post-hackathon.
Only references were in checkpoints/compactions, no callers
in scripts/, prompts/, package.json, or workflows.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Add a "5-minute judge tour" section to README naming the exact
click-path for evaluation: pitch → INDEX → one week's visual review
→ critic-genealogy.ts → optional live --dry-run.

Add demo-output/landing-page/INDEX.md narrating the 11-week LP
timelapse, one beat per week with classification + links to that
week's screenshots and visual review. Turns scattered week directories
into a scannable evidence path.

Drop the "rendered MP4 hosted externally — link in submission form"
references from README, ARCHITECTURE, and FEATURES. The video link
is owned by the submission form, not the repo, and pointing at a
form judges may not have open is dead weight. The per-week assets
plus INDEX.md carry the visual evidence on their own.

Validate green: 176 tests pass, 0 lint warnings, 0 type errors,
0 markdown errors, 19 specs valid, 7 findings files valid.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
@richsak richsak merged commit 4a35def into dev Apr 27, 2026
5 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant