01 · Roasts
The CI split-brain
ZeroCache and courier-rto-fraud-audit automate the boring stuff; ZENITH's 35-star search engine still ships without CI.
Stars chose one child
ZENITH holds 35 of 42 total stars. The rest of the portfolio is shipping hard while adoption remains mostly theoretical.
Horizontal systems gremlin
168 recent commits across analyzed repos, 849 commits this year, and projects ranging from LSM storage to WASM board games: focus is apparently optional.
Bug report with teeth
mastra-issues contains four runnable high-severity reproductions, but its test command exits 1 and CI is absent. Chaos, documented.
Built using
Zoral
Shadows one worker for a week, then takes over their job with zero extra setup. Behaves exactly like the original.
zoral.ai
02 · Category breakdown
- Impact25% weight68C
- Consistency20% weight65C
- Quality20% weight75B
- Depth15% weight60C
- Breadth10% weight80A
- Community10% weight50D
03 · Stats
365-day commit heatmap
140 active days
Language distribution
- Rust33%
- TypeScript30%
- Go23%
- Astro3%
- CSS3%
- HTML2%
- Other6%
04 · Numbers
Owned repos
non-fork
22
Commits
last 12 months
849
Followers
6
Joined GitHub
Jan 2020
05 · Top repos
shramanb113 /
ZeroCache
ZeroCache is a substantial, documented Rust workspace for embedding and LLM-response caching, with provider adapters, Redis/sled backends, semantic search, streaming, a dashboard, demos, tests, CI, and a scratch Docker runtime; adoption remains minimal at 1 star.
shramanb113 /
courier-rto-fraud-audit
A documented, tested Python analytics application with Polars feature engineering, K-Means courier risk clustering, Streamlit visualization, SQL persistence, S3 ingestion, Docker packaging, and CI; adoption remains un demonstrated at 0 stars.
shramanb113 /
Zerops-Swarmforge
SwarmForge is a substantial TypeScript Zerops hackathon system: a Fastify control plane coordinates Architect, Coder, and Deployer agents through NATS, Postgres, Valkey, generated code checks, and deployment manifests.
shramanb113 /
ZENITH
ZENITH is a substantial Go search engine combining hybrid ranking, fuzzy and semantic retrieval with WAL/LSM persistence, a CLI, gRPC service, and embedded ONNX inference; it is well documented and tested but lacks CI, licensing, and broad adoption evidence.
shramanb113 /
9-men-morris
A shipped Nine Men's Morris product with a live playmorris.vercel.app browser UI, typed Rust engine/WASM adapter, React frontend, and unusually careful rule/search implementation, but only 1 star and no CI, license, or repository-level test automation.
shramanb113 /
mastra-issues
A focused, well-documented JavaScript repository containing four high-severity, runnable Mastra bug reproductions with automated checks, but little demonstrated adoption and no formal test suite, CI, license, or typed implementation.
shramanb113 /
shramanb113
A polished GitHub profile repository centered on a detailed README, with named projects and OSS contributions documented, but no source code, tests, CI, or demonstrated adoption for this repo itself.
06 · Timeline
- Jan 19, 2020Joined GitHub
- Jan 24, 2026Created shramanb113 — My github front page
- Feb 1, 2026Created ZENITH — From scratch search engine in Go - no Elasticsearch, no Lucene, just LSM trees, and hybrid ranking
- Jul 8, 2026Created courier-rto-fraud-audit
- Jul 22, 2026Created mastra-issues — All of the mastra issues are to be bundled here
- Jul 23, 2026Created ZeroCache — Cache your embeddings, not your compute bill - Rust-native, provider-pluggable cache for RAG ingestion pipelines.
- Aug 8, 2026Created Zerops-Swarmforge — Multi-Agent engineering team using Zerops
- Aug 14, 2026Created 9-men-morris — Game engine for 9 men morris inspired by POPO
- Sep 2, 2026Most recent push to courier-rto-fraud-audit
07 · Compare
08 · Rubric
How this score was produced
Overall = Σ (category × weight) + gentle top-end curve
Tier thresholds
▸ How the pipeline works
- 01Scrape.Pull every non-fork repo pushed in the last 90 days, plus your contribution calendar, followers, and language byte counts — straight from GitHub's REST & GraphQL APIs.
- 02Triage.A small model reads every repo's file tree + README and picks the 20 files per repo that actually reveal how you code.
- 03Grade each repo. All repos run in parallel through a fast scoring model that reads the picked files and rates each one independently on Impact, Quality, and Depth — with evidence citations.
- 04Aggregate. A larger reasoning model combines the per-repo scores with server-computed stats (heatmap, commit cadence, language entropy, follower count) to produce the 6-dimension profile score + roasts.
- 05Correct.Deterministic server-side checks enforce anchor-scale floors (e.g. a profile with 2,000+ public commits can't score 30 Consistency) and recompute the final verdict.
~90 seconds per profile, ~$0.25 in compute. Total of ~240 files read across your top-12 repos. One rating per GitHub account per day.
▸ Data sources & caveats
- Heatmap & commit totals: GitHub GraphQL
contributionsCollection— covers the last 365 days, includes private repos when the user has opted in (default). - Language %: byte totals across the top 30 owned non-fork repos.
- Curve: a small upward nudge centered on raw score ≈ 70, capping at 100. Prevents specialists from being unfairly penalised for narrow breadth.
- Anchor corrections: when server-measured signals (e.g. privateWorkLikely, multiRepoVolume, follower count) mandate a minimum category score, the aggregation step enforces it. These are signal-conditional, not identity-based floors.