Explore · Ranking › sybil-solutions/inference-optimisation-skill
Agent skill for hill-climbing local LLM serving speed (decode, speculative decoding, prefill, KV, startup, TTFT) on any hardware and quantization
#36best daily rank (Nov 17)
+108stars on Nov 20
+1,384stars in 7 days
1,439total stars
Where it stands
Cooling · 19% of its Nov 17 peak
- Best day: +579 stars on Nov 17.
- Nov 20: 19% of that.
- It fell below half of its best day 1 day later.
- Day 5 of its run (the days in a row with at least a tenth of its best day).
Stars per day
Nov 16peak 579 a dayNov 21 so far
Fake-star check
No red flags
- Only 1 commit (common for new launches)
Checks forks, watchers, issues and how the stars arrived. A flag is a reason to look closer, not proof. How the check works
Why it may be trending
Part of the agent skills wave · 25 on the listBrand new · created Nov 16
Where people talk
X: 17 posts linked it in the last 7 days, reposts not counted. See them on X ↗
On YouTube
- [GitHub Tracer EP.26 · 11.18] 구글 프로모션과 무중단 Rust 전환 #ShortsOrangeCubase · 129 views · Nov 18
- Trending on GitHub right now: sybil-solutions/inference-optimisation-skill. Agent skill for hill-cliEverything For AI · 23 views · Nov 17
- 【inference-optimisation】ローカルLLMを速くするスキルを Claude Code で試した|スキルありは読み込み7%速く、1.4倍はスキルなしの1回で出力が一部変わるさつきのOSS研究室 · 19 views · Nov 18
Videos on YouTube whose description links this repo. YouTube Terms of Service
Days in the ranking
- Nov 20#99
- Nov 19#98
- Nov 18#79
- Nov 17#36
- Nov 16#55
Badge for your README
It shows the best daily rank and updates itself. Paste this into your README:
[](https://truestars.dev/r/sybil-solutions/inference-optimisation-skill/)