01 · Roasts
The 2-Second Architect
role-authorization-reconciler was born and 'completed' in 2 seconds flat. evidence-admissibility-guard took a leisurely 103 seconds. Your entire portfolio is speed-running the concept of 'shipping.'
Copy-Paste Factory
12 repos analyzed, 12 repos follow the exact same pattern: one Python file, embedded fixtures, no tests, no CI, pushed in a single commit. This isn't a portfolio, it's a template with a find-replace on the filename.
task-issuance-collision-guard is Truncated Mid-Function
Your source file literally ends mid-statement: `for t in all_active_ta`. You committed incomplete code, pushed it publicly, and called it done. That's a new kind of courage.
Zero Everything
0 stars, 0 forks, 0 followers, 0 external PRs, 0 issues. The GitHub social graph doesn't know you exist, and based on the heatmap, neither does GitHub for most of the year.
Architecture Without Code
nft-ownership-sweeper-architecture is literally just design docs labeled '0.1.0-draft' with no implementation. You designed a house and submitted the napkin sketch as a construction project.
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% weight15F
- Consistency20% weight55D
- Quality20% weight36F
- Depth15% weight10F
- Breadth10% weight40D
- Community10% weight5F
03 · Stats
365-day commit heatmap
17 active days
Language distribution
- Python81%
- JavaScript15%
- HTML2%
- CSS1%
- Other1%
04 · Numbers
Owned repos
non-fork
25
Commits
last 12 months
36
Followers
0
Joined GitHub
Sep 2020
05 · Top repos
mferdickbutt /
mission-sanction-exposure-reducer
Single-file Python reducer with deterministic risk classification logic. No tests, CI, or license. Created and pushed same day (2026-04-25) with minimal commits. Functional but thin architecture.
mferdickbutt /
role-authorization-reconciler
Single-file Python utility for role authorization reconciliation with README docs, classification logic, and fixture data. One commit in a few seconds; minimal scope; untyped, no tests or CI.
mferdickbutt /
manual-review-prioritizer
Single-file Python utility for ranking manual-review cases by risk and urgency. Untyped, no tests/CI, but self-contained with clear business logic and 13 embedded fixture cases. Created and pushed same day (2026-04-21); minimal commit history.
mferdickbutt /
accepted-task-bypass-router
Single-file Python tool for classifying task authorization bypass patterns. Functional, readable, with embedded fixtures and deterministic JSON output. Created and pushed same day; no tests, CI, license, or typed annotations.
mferdickbutt /
nft-ownership-sweeper-architecture
Architecture-only draft package with 0 stars, no README, no code implementation—just design documentation in ARCHITECTURE.md and related docs. Single commit, 7 KB repo from 2026.
mferdickbutt /
evidence-admissibility-guard
Ultra-fresh one-file Python reducer for evidence admissibility checking; no tests, CI, or git history. Created 2026-04-26, last push same day. Untyped Python with domain-specific logic but bare-minimum project scope.
mferdickbutt /
task-spec-linter
Single-file Python linter with embedded test fixtures and zero dependencies. Created and pushed same day (2026-04-19); one commit. No tests, no CI, no license. Untyped Python with functional scope-checking logic but minimal production readiness.
mferdickbutt /
task-generation-suppression-reducer
A single-file Python utility (4KB) with clear type hints and docstrings that classifies task-generation requests. Created and pushed same day (Apr 15, 2026), no tests/CI/license. Well-documented in README but sparse codebase; non-trivial logic but no external adoption signals.
mferdickbutt /
policy-diff-reporter
Single-file authorization policy diff reporter with embedded test scenarios. No tests, no CI, no types, no license. Uploaded moments ago (1 commit in 2 seconds).
mferdickbutt /
task-issuance-collision-guard
One-file Python utility (6 KB) implementing a collision guard reducer for task issuance evaluation. Created and pushed same day (2026-04-26), untyped, no tests or CI, with embedded fixtures and self-test assertions but incomplete implementation (source file truncated mid-function).
mferdickbutt /
postfiat-ipfs-audit-v2
One-off IPFS latency audit tooling (5 Python scripts) with hardcoded configs, 0 stars, single-day creation. Lacks README, tests, CI, and documented architecture. Non-trivial technical execution (curl+check-host.net integration, Markdown report generation) but no reusability, maintenance, or broader impact.
mferdickbutt /
refusal-cluster-suppression-reducer
One-shot script dump with embedded fixtures. Zero stars, created and pushed same minute (2026-04-24T14:51:40–42Z). Single file, minimal real utility; no tests, CI, or typed annotations. Reads embedded RECORDS only; zero external adoption or production use.
06 · Timeline
- Sep 10, 2020Joined GitHub
- Apr 15, 2026Created policy-diff-reporter — Cross-surface authorization policy diff reporter — compares two policy versions against a shared scenario corpus and surfaces decision flips with risk-tiered operator handoff queue
- Apr 15, 2026Created task-generation-suppression-reducer — Deterministic task-generation suppression reducer that classifies candidate requests against recent fingerprints, refusal history, authorization state, and load signals
- Apr 16, 2026Created nft-ownership-sweeper-architecture — NFT Ownership Sweeper — NestJS Architecture Package
- Apr 19, 2026Created task-spec-linter — Pre-issuance task spec linter for draft network task quality gates. Zero-dependency Python.
- Apr 20, 2026Created postfiat-ipfs-audit-v2 — PostFiat public NFT/IPFS latency audit (Fresh Slice v2): Singapore+Sao Paulo+Vancouver x 5 gateways x 5 fresh CIDs
- Apr 21, 2026Created accepted-task-bypass-router — Accepted-task bypass root-cause router — self-contained Python reducer for auditing assignment-gate leaks
- Apr 21, 2026Created manual-review-prioritizer — Deterministic Python reducer that ranks manual-review cases by blocked PFT, authorization risk, queue age, and evidence gaps
- Apr 23, 2026Created role-authorization-reconciler — Deterministic Python reducer that reconciles project role assignments against an authorization source of truth and emits a prioritized cleanup queue
- Apr 24, 2026Created refusal-cluster-suppression-reducer — Groups refused/cancelled network tasks into root-cause clusters, scores reissue risk, and recommends suppression/rewrite/merge/reissue actions
- Apr 25, 2026Created mission-sanction-exposure-reducer — Self-contained Python reducer for mission-level sanction risk summaries with deterministic JSON output
- Apr 26, 2026Created task-issuance-collision-guard — Deterministic collision guard reducer that evaluates proposed task issuances against backlog, owner load, refusal patterns, and scope fingerprints
- Apr 26, 2026Created evidence-admissibility-guard — Deterministic reducer that checks network-task evidence admissibility before reviewer assignment
- Apr 26, 2026Most recent push to evidence-admissibility-guard
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.