GitHub Trending daily report · 2026-09-20
Period and coverage
2026-09-20 to 2026-09-20 · Asia/Shanghai. 1 saved daily snapshots, 15 unique repositories and 15 repository observations.
The recorded daily Star totals add up to 7,848. For multi-day periods this is the sum of the saved daily signals, not a measurement of net Stars gained between two exact timestamps. Only available snapshots are included.
Leading recorded signals
-
cloudflare/security-audit-skill — 3,162 recorded daily Stars, best rank #1, present in 1 saved snapshot. A coding-agent skill for multi-phase security audits with independently verified, machine-readable findings
-
trycua/cua — 1,124 recorded daily Stars, best rank #2, present in 1 saved snapshot. Scale computer-use 2.0 with open-source drivers, cross-OS fleets, and benchmarks for training, evaluation, and data generation.
-
addyosmani/agent-skills — 547 recorded daily Stars, best rank #3, present in 1 saved snapshot. Production-grade engineering skills for AI coding agents.
All projects in this period
| Project | Recorded daily Stars | Best rank | Snapshots |
|---|---|---|---|
| cloudflare/security-audit-skill | 3,162 | 1 | 1 |
| trycua/cua | 1,124 | 2 | 1 |
| addyosmani/agent-skills | 547 | 3 | 1 |
| anthropics/claude-code | 482 | 5 | 1 |
| Open-Dev-Society/OpenStock | 477 | 6 | 1 |
| asciimoo/hister | 430 | 10 | 1 |
| coder/coder | 406 | 4 | 1 |
| higgsfield-ai/higgsfield | 314 | 7 | 1 |
| anthropics/knowledge-work-plugins | 280 | 13 | 1 |
| cactus-compute/needle | 207 | 14 | 1 |
| ruanyf/weekly | 151 | 11 | 1 |
| docling-project/docling | 94 | 8 | 1 |
| cloudflare/quiche | 84 | 9 | 1 |
| yynxxxxx/Codex-X | 59 | 15 | 1 |
| ZuodaoTech/everyone-can-use-english | 31 | 12 | 1 |
Daily records
- 2026-09-20 — 15 projects, 7,848 daily Stars.
Reading the signals
Star counts and rank changes measure attention in the saved Trending observations, not software quality, security or suitability. A gap in this archive does not establish that a project left GitHub Trending. Follow the project details and official documentation to check requirements before choosing a tool.
This English report is generated from the same saved snapshots and reporting period as the Chinese edition. It is a data summary; the Chinese edition may also contain editorial interpretation. No missing observations or publication timestamps are invented.