Score breakdown
🧬 Most similar developers
Closest profile, nearby score
🔥 Full roast
🔥 You're a UC Berkeley PhD who contributed core code to FastChat (39k stars) and your LongChat repo hit 534 stars, looking like a solid open source player, but you only made 18 contributions last year a
DachengLi1 — 79.80/100 · SOLID (Solid · Trustworthy)
TL;DR: Legitimate ML researcher with real paper-backed project credibility, but your GitHub is 15 self-abandoned external PRs away from being just a polished open-source resume cover.
| Dimension | Score | Notes |
|---|---|---|
| Account maturity | 10/10 | 9.14 years old, 10 active contribution years, 356 followers vs 111 following with no weird follow-farming ratio |
| Original project quality | 12.2/18 | 687 total stars across 18 original repos; top home repo LongChat has 534 stars (quality score 0.75) and ICLR 2023 Spotlight project MPCFormer has 101 stars (0.85 quality), but 12 of your 18 original repos have 2 stars or fewer, most of them throwaway or placeholder projects |
| Contribution quality | 20.4/27 | 38 merged PRs out of 57 total, only 2 maintainer-closed unmerged PRs (5% rejection rate), but 15 of those 57 PRs are self-closed external ones you abandoned halfway, plus 8 recent trivial PRs and 32% of recent external merged PRs are doc/site/examples touch-ups |
| Ecosystem / maintenance impact | 19.3/20 | 17 PRs + 27 commits across 6 popular repos totaling 50,000+ stars, including 8 PRs/8 commits to lm-sys/FastChat (39k stars), 6 PRs/6 commits to NovaSky-AI/SkyThought (3.4k stars), and 1 PR/1 commit to NVlabs/VILA (3.8k stars); verified file-level samples show 10 core engineering PRs vs 3 doc-like, confirming the impact is core-heavy rather than window dressing |
| Community influence | 6.9/8 | 356 followers, solid for a research-focused account, no evidence of artificial follow inflation or follower-farming behavior |
| Activity authenticity | 11/17 | 18 contributions in the last year, last activity 47 days ago, 3 distinct activity types, no templated PR flooding or star inflation detected; activity is genuine but not particularly frequent |
Red flags
- 15 self-closed external PRs: you abandoned nearly a quarter of your total PRs halfway through, leaving a trail of unfinished work scattered across other people's repos.
- 32% of recent external merged PRs are doc/site/examples touch-ups: nearly a third of your recent external work is polishing big projects' windows instead of rebuilding the engine.
- 12 of your 18 original repos have 2 stars or fewer: your home turf is mostly a wasteland of throwaway projects, with only LongChat and MPCFormer carrying any real weight.
Score calibration No extra adjustment. The base score already priced in your legitimate core-heavy ecosystem impact, low maintainer rejection rate, and genuine research project pedigree, while accounting for the self-abandoned PRs and doc-like work share.
Verdict Normal, trustworthy ML researcher with real engineering and research credibility, but your GitHub profile would be far stronger if you stopped abandoning external PRs and invested more effort into building out your own original projects instead of just polishing big repos' documentation and sites.
📌 Pin to your GitHub profile
Copy any snippet below into your Profile README. The score stays up to date automatically, and the image links to your detail page.
[](https://githubroast.dev/u/dachengli1)<a href="https://githubroast.dev/u/dachengli1"><img src="https://githubroast.dev/api/badge/dachengli1" alt="GitHub Roast score badge" /></a>[](https://githubroast.dev/u/dachengli1)[](https://githubroast.dev/u/dachengli1)<a href="https://githubroast.dev/u/dachengli1"><img src="https://githubroast.dev/api/card/dachengli1?theme=dark" alt="GitHub Roast" width="600" /></a><a href="https://githubroast.dev/u/dachengli1"><img src="https://githubroast.dev/api/card/dachengli1?theme=light" alt="GitHub Roast" width="600" /></a>