Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Generates markdown digests and CSV exports for GitHub initiative health. Use when reporting on issue/PR progress across a milestone or project.
.claude/skills/athola-github-initiative-pulse/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 17% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 2% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -73% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -68% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -47% | 0% |
minister:dora-metrics)minister:release-health-gates)Turns tracker data and GitHub board metadata into initiative-level summaries. Provides markdown helpers and CSV exports for pasting into issues, PRs, or Discussions.
tracker.py add or sync from GitHub Projects.tracker.py status --github-comment or module snippets.| Metric | Description | |--------|-------------| | Completion % | Done tasks / total tasks per initiative. | | Avg Task % | Mean completion percent for all in-flight tasks. | | Burn Rate | Hours burned per week (auto-calculated). | | Risk Hotlist | Tasks flagged priority=High or due date in past. |
phase in the tracker record.If metrics appear outdated, ensure tracker.py has successfully synced with GitHub. If the Markdown digest renders incorrectly in GitHub, check for unescaped characters in task titles or missing newlines between table rows.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 9,692 | 8,462 | -13% | 1 | 1 | 0% | 1,547 | 1,816 | +17% | 0 | 0 | — |
case-02 | fail→fail | 14,124 | 8,847 | -37% | 1 | 1 | 0% | 2,484 | 1,888 | -24% | 0 | 0 | — |
case-03 | fail→pass | 9,946 | 7,441 | -25% | 1 | 1 | 0% | 1,664 | 1,689 | +2% | 0 | 0 | — |
case-04 | fail→pass | 16,699 | 2,317 | -86% | 1 | 1 | 0% | 2,879 | 773 | -73% | 0 | 0 | — |
case-05 | fail→pass | 16,368 | 2,834 | -83% | 1 | 1 | 0% | 2,561 | 817 | -68% | 0 | 0 | — |
case-06 | fail→pass | 12,237 | 4,607 | -62% | 1 | 1 | 0% | 2,069 | 1,102 | -47% | 0 | 0 | — |
case-07 | pass→pass | 9,230 | 1,753 | -81% | 1 | 1 | 0% | 1,509 | 602 | -60% | 0 | 0 | — |
case-08 | fail→pass | 11,485 | 2,486 | -78% | 1 | 1 | 0% | 1,958 | 736 | -62% | 0 | 0 | — |
case-09 | pass→pass | 12,983 | 2,532 | -80% | 1 | 1 | 0% | 2,199 | 792 | -64% | 0 | 0 | — |
case-10 | pass→pass | 10,717 | 1,874 | -83% | 1 | 1 | 0% | 1,677 | 613 | -63% | 0 | 0 | — |
case-11 | fail→pass | 7,500 | 1,604 | -79% | 1 | 1 | 0% | 1,344 | 587 | -56% | 0 | 0 | — |
case-12 | fail→pass | 10,842 | 2,368 | -78% | 1 | 1 | 0% | 1,827 | 699 | -62% | 0 | 0 | — |
case-13 | fail→pass | 9,595 | 1,450 | -85% | 1 | 1 | 0% | 1,510 | 593 | -61% | 0 | 0 | — |
case-14 | fail→pass | 12,673 | 3,269 | -74% | 1 | 1 | 0% | 2,118 | 898 | -58% | 0 | 0 | — |
case-15 | pass→pass | 5,199 | 1,428 | -73% | 1 | 1 | 0% | 791 | 528 | -33% | 0 | 0 | — |
case-16 | fail→pass | 7,224 | 1,595 | -78% | 1 | 1 | 0% | 1,043 | 632 | -39% | 0 | 0 | — |
case-17 | fail→pass | 10,197 | 1,596 | -84% | 1 | 1 | 0% | 1,569 | 626 | -60% | 0 | 0 | — |
case-18 | fail→pass | 6,259 | 1,656 | -74% | 1 | 1 | 0% | 821 | 600 | -27% | 0 | 0 | — |
case-19 | fail→pass | 7,703 | 1,981 | -74% | 1 | 1 | 0% | 1,113 | 659 | -41% | 0 | 0 | — |
case-20 | fail→pass | 6,024 | 2,723 | -55% | 1 | 1 | 0% | 950 | 784 | -17% | 0 | 0 | — |
case-21 | fail→pass | 12,833 | 1,788 | -86% | 1 | 1 | 0% | 1,907 | 643 | -66% | 0 | 0 | — |
case-22 | pass→pass | 9,033 | 1,952 | -78% | 1 | 1 | 0% | 1,306 | 656 | -50% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +73 percentage points is the difference between those two pass rates over the 22 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.