⛏ Clark Fork Gold — Dev Activity

Ownership shares are calculated from combined totals across this repository and pre-migration history.
Formula: 60% True Value (code output)  ·  20% Commits  ·  20% Pull Requests
These percentages represent fair contribution shares to be used for future income distribution.

Source repo: barbrickdesign/barbrickdesign.github.io   ·   Period: 2026-04-01 – 2026-04-06   ·   This represents development work done in the source repository before the code was migrated into this collaborative repo.
Per-commit diff data is unavailable for this period; values are estimated using the same methodology constants and an assumed average of 200 lines/commit. The True (No-AI) value estimates are included in the Contribution Ownership calculation above.

AI-Assisted Value (what you see as “AI Hours” / “AI Value”)
Every commit represents two real phases:
  • Human prompting — 5 min (constant: writing the prompt, reviewing output)
  • AI execution — scales with the commit’s actual diff size (lines added + removed):
       < 20 lines → 10 min  |  20–99 → 15 min  |  100–299 → 20 min  |  300–499 → 25 min  |  500+ → 30 min

True Value (No-AI) — what the same work would cost without AI assistance
  • Typing speed — 40 WPM × 5 chars/word = 200 chars/min
  • Avg line length — 35 chars → 0.175 min/line raw typing
  • Overhead multiplier — 14× (thinking 4× · debugging 2× · testing 1.5× · planning 1.25×)
  • Effective rate — 2.45 min/line without AI (≈ 24 lines/hour)
  • Floor — 15 min minimum per commit (planning/setup overhead)

Anti-Gaming Quality Scores (0–100)
Each commit is scored for contribution quality. Flags reduce the score:
  ⚠ generic_message −25 pts — vague commit message with no specifics
  📎 binary_only −30 pts — all changed files are binary/non-code
  ⚡ mass_dump −20 pts — huge diff paired with a short/generic message
  □ empty_commit −10 pts — no lines changed and no tracked files
  ♻ duplicate_spam −15 pts — same message used 3+ times by the same developer

Ownership Share Methodology
Ownership % = weighted blend of three normalized contribution dimensions:
  • True Value (60%) — code output quality × quantity (lines written without AI)
  • Commits (20%) — current-repo commits + legacy commits combined
  • Pull Requests (20%) — current-repo PRs + legacy PRs combined
Each dimension is normalized to [0,1] before weighting so no single metric dominates.

Hourly rate — $75 / hr (applied to both AI-assisted and no-AI estimates)
Work categories and file types are detected automatically from commit messages and file extensions.