v0.98.5

Changes

• Fast profile entry: close RoastAccessModal immediately on generate, show profile with shimmer skeleton loading states
• Roasting skeletons: roast text, hashtags, hook, and tldr all show polished shimmer animations while AI generates
• Animation replay: ProfileResult remounts on entry so entrance animations play fresh
• Witty hashtag pools: 13 evidence-conditioned tag pools for local fallback (ship_fast, build_beast, debug_hell, etc.)
• Hashtag limit: 8→10
• AI prompt: bolder hashtag instructions, ZH tag limit 16→24 chars

v0.98.4

v0.98.4

Local-first AI coding personality profiler, usage dashboard, and evidence-grounded roast.

Highlights

• 16-type vibe personality engine with illustrated characters
• Local-only TokenTracker integration for evidence-grounded roasts
• Bilingual support: English + 中文 README
• Dashboard with share-poster modal and i18n
• CLI with inspect and serve commands

Install

npm install -g vibe-roast

Recent changes

• docs: add Chinese README and trim AI-speak from both versions
• Keep TokenTracker integration local-only
• Bundle TokenTracker token collection