Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →User-level AI agents that travel with you across projects
.claude/skills/github-personal-squad/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -20% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -56% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -9% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -49% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -11% | 0% |
A personal squad is a user-level collection of AI agents that travel with you across projects. Unlike project agents (defined in a project's .squad/ directory), personal agents live in your global config directory and are automatically discovered when you start a squad session.
~/Library/Application Support/squad/personal-squad/ # macOS
~/.config/squad/personal-squad/ # Linux
%APPDATA%/squad/personal-squad/ # Windows
├── agents/
│ ├── {agent-name}/
│ │ ├── charter.md
│ │ └── history.md
│ └── ...
└── config.json # Optional: personal squad configSQUAD_NO_PERSONAL=1 to disable ambient discoverysquad personal init — Bootstrap a personal squad directorysquad personal list — List your personal agentssquad personal add {name} --role {role} — Add a personal agentsquad personal remove {name} — Remove a personal agentsquad cast — Show the current session cast (project + personal)See templates/ghost-protocol.md for the full rules. Key points:
.squad/ stateOptional config.json in the personal squad directory:
json{ "defaultModel": "auto", "ghostProtocol": true, "agents": {} }
SQUAD_NO_PERSONAL — Set to any value to disable personal squad discoverySQUAD_PERSONAL_DIR — Override the default personal squad directory path| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 31,220 | 6,513 | -79% | 1 | 1 | 0% | 1,705 | 1,358 | -20% | 0 | 0 | — |
case-02 | fail→pass | 17,023 | 4,869 | -71% | 1 | 1 | 0% | 2,803 | 1,227 | -56% | 0 | 0 | — |
case-03 | fail→pass | 13,237 | 8,692 | -34% | 1 | 1 | 0% | 1,934 | 1,758 | -9% | 0 | 0 | — |
case-04 | fail→pass | 19,621 | 3,398 | -83% | 1 | 1 | 0% | 1,520 | 778 | -49% | 0 | 0 | — |
case-05 | fail→pass | 6,495 | 4,204 | -35% | 1 | 1 | 0% | 902 | 806 | -11% | 0 | 0 | — |
case-06 | fail→pass | 16,340 | 2,551 | -84% | 1 | 1 | 0% | 2,595 | 800 | -69% | 0 | 0 | — |
case-07 | fail→pass | 17,136 | 2,378 | -86% | 1 | 1 | 0% | 1,174 | 782 | -33% | 0 | 0 | — |
case-08 | pass→pass | 24,294 | 3,475 | -86% | 1 | 1 | 0% | 3,960 | 844 | -79% | 0 | 0 | — |
case-09 | fail→pass | 41,223 | 2,868 | -93% | 1 | 1 | 0% | 3,822 | 858 | -78% | 0 | 0 | — |
case-10 | fail→pass | 17,590 | 3,564 | -80% | 1 | 1 | 0% | 2,981 | 927 | -69% | 0 | 0 | — |
case-11 | fail→pass | 32,580 | 2,618 | -92% | 1 | 1 | 0% | 3,451 | 841 | -76% | 0 | 0 | — |
case-12 | pass→pass | 17,245 | 3,140 | -82% | 1 | 1 | 0% | 2,652 | 912 | -66% | 0 | 0 | — |
case-13 | pass→pass | 17,306 | 7,043 | -59% | 1 | 1 | 0% | 2,370 | 1,468 | -38% | 0 | 0 | — |
case-14 | pass→pass | 9,869 | 4,081 | -59% | 1 | 1 | 0% | 1,249 | 887 | -29% | 0 | 0 | — |
case-15 | fail→pass | 6,172 | 2,947 | -52% | 1 | 1 | 0% | 875 | 824 | -6% | 0 | 0 | — |
case-16 | fail→pass | 11,413 | 2,951 | -74% | 1 | 1 | 0% | 1,619 | 809 | -50% | 0 | 0 | — |
case-17 | pass→pass | 17,519 | 10,601 | -39% | 1 | 1 | 0% | 2,733 | 2,251 | -18% | 0 | 0 | — |
case-18 | pass→pass | 25,108 | 2,975 | -88% | 1 | 1 | 0% | 2,709 | 727 | -73% | 0 | 0 | — |
case-19 | pass→pass | 12,681 | 4,171 | -67% | 1 | 1 | 0% | 1,770 | 863 | -51% | 0 | 0 | — |
case-20 | fail→pass | 10,507 | 3,469 | -67% | 1 | 1 | 0% | 1,578 | 776 | -51% | 0 | 0 | — |
case-21 | pass→pass | 15,989 | 9,973 | -38% | 1 | 1 | 0% | 2,389 | 1,795 | -25% | 0 | 0 | — |
case-22 | pass→pass | 14,518 | 24,262 | +67% | 1 | 1 | 0% | 2,187 | 2,539 | +16% | 0 | 0 | — |
case-23 | pass→pass | 20,334 | 17,592 | -13% | 1 | 1 | 0% | 3,179 | 3,013 | -5% | 0 | 0 | — |
case-24 | pass→pass | 10,948 | 11,391 | +4% | 1 | 1 | 0% | 2,020 | 2,487 | +23% | 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. 24 cases were attempted. The headline lift of +54 percentage points is the difference between those two pass rates over the 24 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.