lauytgary/dsh_hud_plugin0

dsh-stats-hud

Sci-fi HUD: turns the session stats line into game-style level bars, a speedometer and a rolling token counter.

包名
dsh-stats-hud
版本
0.1.0
许可证
MIT
最近更新
2026年8月18日

安装

$npx -p @deepseek-ai/dsh dsh plugin --profile web add github:lauytgary/dsh_hud_plugin

English · 简体中文

dsh-stats-hud

A sci-fi HUD for DeepSeek Harness: the session's stats become game-style level bars in a vertical column fixed to the far right edge of the web UI — without touching the original stats line.

screenshot

Instruments (all-English, LLM terminology)

InstrumentDataFull scalePast full scale
CLOCK badgeLocal time (24h) + DS API PEAK / DS API OFF PEAK ratePeak = Beijing 09:00-12:00 / 14:00-18:00 (auto-converted from local time)PEAK orange tint, OFF-PEAK green tint
STEPS / TURN rolling rowsSteps / turns as odometer drums (like the CONTEXT rows)Drums spin up on mount, roll on change
LLM / TOOLS dual barTwo columns (labels over values), bar segments = raw LLM:TOOLS time ratioNo cap — 2:1 time means 2:1 bar
THROUGHPUT gaugetokens/s (throughput), centered title, combined centered readout (146 tok/s)Redline auto-scales 200→300→400… (arc ticks follow)
CONTEXT USAGE barContext-window usage % with 3 segments: Sys Prompt (gray) / Tools (blue) / Messages (purple) by token ratio0-100%≥80% whole bar turns solid red; hover shows the three token counts
CACHE HIT barCache-hit %0-100%= 540 && bjMin = 840 && redline) redline += 100` auto-scales)
  • ContextUsageBar: segment colors and the ≥80% solid-red threshold; the hover tooltip reads systemTokens / toolsTokens / messageTokens from the contextBreakdown projection
  • Rolling counter: DRUM (3× 0-9), DRUM_H = 15 (px per digit), RollingValue's carry/borrow formula and mount spin-up
  • CSS: position:fixed; right:12px; tiers in tierOf(space) (≥200 full / ≥90 mini / else hidden) and the .gsh-root.gsh-* rules

Publishing to npm (optional)

# remove "private": true from package.json, then
npm publish
# users install with:
dsh plugin --profile web add dsh-stats-hud

License

MIT