Skip to content

Improve clone-website skill fidelity guardrails#19

Open
YurunChen wants to merge 1 commit into
aiming-lab:mainfrom
YurunChen:feat/clone-skill-fidelity
Open

Improve clone-website skill fidelity guardrails#19
YurunChen wants to merge 1 commit into
aiming-lab:mainfrom
YurunChen:feat/clone-skill-fidelity

Conversation

@YurunChen
Copy link
Copy Markdown

Summary

  • upgrades .claude/skills/clone-website/SKILL.md with a snapshot contract requirement to keep upstream baselines stable
  • adds entity-level resource alignment rules (title, upstream_url, image_url, local_path) to reduce title/image/link drift
  • adds interactive parity guidance and hard acceptance gates (broken-image, semantic, visual, and restart determinism checks)

Why

Recent mirror alignment work exposed recurring regressions where pages looked structurally correct but drifted semantically (mismatched image/title/link, non-functional visual-only chips/tabs, and weak fidelity acceptance). This update makes those checks explicit in the phase-1 clone workflow.

Validation

  • reviewed the updated skill section flow to ensure it remains consistent with existing clone workflow steps
  • verified commands/path conventions remain aligned with current repository patterns

Made with Cursor

Adds snapshot contract requirements, entity-level asset-text binding rules, interactive parity guidance, and hard acceptance gates so new mirrors converge to higher visual fidelity with fewer semantic alignment regressions.
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