01 · Roasts
Commit furnace
1,252 commits this year and a level-4 heatmap run, yet 22 total stars: the shipping engine is outrunning the audience.
CI knows the assignment
Cognitive's 8-job pipeline tests accessibility and smoke paths; several other substantial repos still show up without CI or tests.
Crate collector
Helios Agent has 12 Rust workspace members built in a one-day burst. Architecture first, calendar later.
Empty-seat reservation
Three scored repositories are empty scaffolds: ESP32-microcontroller-server-uptime-monitor, O-brain, and excalidraw-animate.
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% weight62C
- Consistency20% weight80A
- Quality20% weight75B
- Depth15% weight58D
- Breadth10% weight80A
- Community10% weight80A
03 · Stats
365-day commit heatmap
296 active days
Language distribution
- TypeScript59%
- Rust14%
- MDX10%
- Python7%
- HTML3%
- JavaScript3%
- Other4%
04 · Numbers
Owned repos
non-fork
64
Commits
last 12 months
1,252
Followers
91
Joined GitHub
Jun 2024
05 · Top repos
FirePheonix /
vengeance-ui-blog-template
A documented, typed Next.js blog template with Obsidian bidirectional sync, Markdown rendering, generated OG cards, Docker deployment, tests, and CI; adoption remains limited at 3 stars and 3 forks.
FirePheonix /
shubham-singh-blogs
A substantial typed Next.js blog with searchable markdown content, an AI-readable route, audiobook UI, unit tests, and three CI jobs; it is technically polished but has no demonstrated adoption and no README or alternate documentation.
FirePheonix /
cognitive
A substantial typed Next.js AI knowledge/docs site with interactive search, notebooks, audio, WebGL books, and a notably comprehensive validation pipeline, but only 1 star and no demonstrated external adoption.
FirePheonix /
far-away-2026
Clawvio is a substantial typed multi-component assistant spanning Next.js, Express/Inngest workflows, Supabase persistence, integrations, and Rust desktop speech recognition, but has only 1 star and no visible tests, CI, or license.
FirePheonix /
helios-agent
A documented 12-crate Rust AI-agent workspace with provider abstraction, SQLite persistence, tools, MCP, memory, skills, cron, and gateway modules; substantial engineering is present, but the repo is a one-day, zero-star project without CI, license, or externally demonstrated adoption.
FirePheonix /
lil-local-mascot
A substantial, documented Rust/egui Windows desktop mascot with activity sensing, SQLite tracking, configurable reminders, animation, tray integration, and optional local LLM speech, but currently has no visible adoption, tests, CI, or license.
FirePheonix /
kokoro-deployed-tts
A documented Rust deployment of Kokoro TTS with an OpenAI-compatible Axum API, Docker/Railway packaging, streaming formats, voice mapping, and timestamp support; adoption is unproven at 0 stars and the snapshot lacks tests and a license.
FirePheonix /
parakeet-tdt-v3-CPU-optimized-android
A typed Kotlin Android speech-to-text app with foreground recording, Parakeet model download/extraction, volume-key accessibility control, and automated APK release, but no visible adoption, documentation, tests, or license.
FirePheonix /
kyrocks
A documented Kyverno/ArgoCD Kubernetes demonstration centered on the require-requests-limits policy, with AWS k3s setup and screenshot-backed validation, but no visible adoption or engineering automation.
FirePheonix /
ESP32-microcontroller-server-uptime-monitor
Empty repository with no source files, documentation, configuration, tests, or recorded commit activity.
FirePheonix /
excalidraw-animate
Empty repository scaffold with no files, documentation, implementation, tests, or recorded activity.
FirePheonix /
O-brain
Empty repository scaffold with no source files, documentation, metadata, or observable development activity.
06 · Timeline
- Jun 27, 2024Joined GitHub
- Jun 9, 2026Created far-away-2026
- Jul 24, 2026Created vengeance-ui-blog-template
- Jul 25, 2026Created shubham-singh-blogs
- Jul 29, 2026Created cognitive
- Aug 1, 2026Created kokoro-deployed-tts
- Aug 2, 2026Created excalidraw-animate
- Aug 2, 2026Created parakeet-tdt-v3-CPU-optimized-android
- Aug 7, 2026Created lil-local-mascot — Tiny animated desktop mascot companion for Windows — watches what you do, reminds you to take breaks, AI speech via Ollama
- Aug 16, 2026Created O-brain
- Aug 16, 2026Created helios-agent
- Aug 18, 2026Created kyrocks
- Aug 29, 2026Created ESP32-microcontroller-server-uptime-monitor
- Sep 2, 2026Most recent push to cognitive
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.