Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Task lifecycle state machine — states, transitions, and task tree model. Use when you need to understand task states, dispatch subtasks, or review completions. Do NOT use for domain-specific procedures — this only covers the universal state machine.
.claude/skills/1mancompany-task-lifecycle/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -22% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -6% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -15% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -63% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -58% | 0% |
Every task in the system follows this state machine:
| State | Meaning | |-------|---------| | pending | Created, waiting to be processed | | processing | Actively being executed by an employee | | holding | Waiting for subtasks to complete or external input | | completed | Employee finished execution, awaiting supervisor review | | accepted | Supervisor approved the deliverable | | finished | Fully done, archived | | failed | Execution failed or supervisor rejected | | blocked | Dependency task failed, cannot proceed | | cancelled | Cancelled |
State flow: pending → processing → completed → accepted → finished ↕ holding (pause/resume) completed → failed (rejection) → processing (retry)
Key distinctions:
Task tree model:
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 9,266 | 6,613 | -29% | 1 | 1 | 0% | 1,534 | 1,199 | -22% | 0 | 0 | — |
case-02 | fail→pass | 13,158 | 9,826 | -25% | 1 | 1 | 0% | 2,306 | 2,158 | -6% | 0 | 0 | — |
case-03 | fail→pass | 11,503 | 7,380 | -36% | 1 | 1 | 0% | 1,867 | 1,590 | -15% | 0 | 0 | — |
case-04 | fail→pass | 10,744 | 2,420 | -77% | 1 | 1 | 0% | 1,966 | 720 | -63% | 0 | 0 | — |
case-05 | pass→pass | 7,244 | 1,998 | -72% | 1 | 1 | 0% | 1,200 | 521 | -57% | 0 | 0 | — |
case-06 | fail→pass | 11,515 | 2,696 | -77% | 1 | 1 | 0% | 1,817 | 769 | -58% | 0 | 0 | — |
case-07 | pass→pass | 7,244 | 2,254 | -69% | 1 | 1 | 0% | 1,235 | 639 | -48% | 0 | 0 | — |
case-08 | fail→pass | 12,214 | 2,080 | -83% | 1 | 1 | 0% | 818 | 539 | -34% | 0 | 0 | — |
case-09 | fail→pass | 6,694 | 2,570 | -62% | 1 | 1 | 0% | 1,043 | 689 | -34% | 0 | 0 | — |
case-10 | pass→pass | 10,057 | 5,069 | -50% | 1 | 1 | 0% | 1,644 | 1,067 | -35% | 0 | 0 | — |
case-11 | fail→pass | 9,558 | 2,306 | -76% | 1 | 1 | 0% | 1,381 | 620 | -55% | 0 | 0 | — |
case-12 | pass→pass | 6,832 | 2,460 | -64% | 1 | 1 | 0% | 1,099 | 631 | -43% | 0 | 0 | — |
case-13 | fail→pass | 13,767 | 1,899 | -86% | 1 | 1 | 0% | 2,231 | 619 | -72% | 0 | 0 | — |
case-14 | pass→pass | 10,740 | 2,979 | -72% | 1 | 1 | 0% | 1,634 | 772 | -53% | 0 | 0 | — |
case-15 | pass→pass | 7,386 | 2,895 | -61% | 1 | 1 | 0% | 1,356 | 695 | -49% | 0 | 0 | — |
case-16 | pass→pass | 3,379 | 1,815 | -46% | 1 | 1 | 0% | 513 | 539 | +5% | 0 | 0 | — |
case-17 | fail→pass | 7,496 | 2,829 | -62% | 1 | 1 | 0% | 1,232 | 751 | -39% | 0 | 0 | — |
case-18 | fail→pass | 8,495 | 1,936 | -77% | 1 | 1 | 0% | 1,359 | 574 | -58% | 0 | 0 | — |
case-19 | pass→pass | 9,509 | 1,852 | -81% | 1 | 1 | 0% | 1,717 | 517 | -70% | 0 | 0 | — |
case-20 | pass→pass | 2,909 | 2,014 | -31% | 1 | 1 | 0% | 521 | 589 | +13% | 0 | 0 | — |
case-21 | pass→pass | 6,795 | 4,164 | -39% | 1 | 1 | 0% | 1,162 | 1,027 | -12% | 0 | 0 | — |
case-22 | pass→pass | 3,103 | 2,071 | -33% | 1 | 1 | 0% | 509 | 611 | +20% | 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, and 21 counted toward the lift figure. The other 1 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +50 percentage points is the difference between those two pass rates over the 21 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.