Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Delegates tasks to Gemini CLI implementing delegation-core for Google's models. Use when delegation-core selects Gemini or 1M+ token context is needed.
.claude/skills/athola-gemini-delegation/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-03 | ✗→✓ | ▲ Improved | -47% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -14% | 0% |
| case-13 | ✗→✓ | ▲ Improved | -33% | 0% |
| case-14 | ✗→✓ | ▲ Improved | -20% | 0% |
| case-15 | ✗→✓ | ▲ Improved | -49% | 0% |
This skill implements conjure:delegation-core for the Gemini CLI. It provides Gemini-specific authentication, quota management, and command construction.
For shared delegation patterns, see Skill(conjure:delegation-core).
Skill(conjure:delegation-core) determines Gemini is suitablegemini CLI is installed and authenticatedconjure:delegation-core)conjure:qwen-delegation)Installation:
bash# Verify installation gemini --version # Check authentication gemini auth status # Login if needed gemini auth login # Or set API key export GEMINI_API_KEY="your-key"
Verification: Run the command with --help flag to verify availability.
bash# File analysis gemini -p "@path/to/file Analyze this code" # Multiple files gemini -p "@src/**/*.py Summarize these files" # With specific model gemini --model gemini-3-pro -p "..." # JSON output gemini --output-format json -p "..."
bashgemini -p "..." > delegations/gemini/$(date +%Y%m%d_%H%M%S).md
For Gemini-specific models, CLI options, cost reference, and troubleshooting, see modules/gemini-specifics.md.
gemini --version and gemini auth status both exit 0 before any task is delegated;missing installation or failed authentication is reported and stops execution.
delegations/gemini/YYYYMMDD_HHMMSS.md (timestamp format matching the Quick Start example), and that file exists on disk after the delegation completes.
conjure:delegation-core selected a different provider (Qwen or local), this skillis not invoked; Gemini delegation only runs when delegation-core explicitly routes to Gemini.
the estimated token count and confirms it falls within Gemini's supported window.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 5,766 | 7,368 | +28% | 1 | 1 | 0% | 370 | 1,212 | +228% | 0 | 0 | — |
case-02 | fail→fail | 20,349 | 7,845 | -61% | 1 | 1 | 0% | 3,463 | 970 | -72% | 0 | 0 | — |
case-03 | fail→pass | 16,192 | 5,263 | -67% | 1 | 1 | 0% | 2,992 | 1,588 | -47% | 0 | 0 | — |
case-04 | fail→fail | 7,219 | 9,042 | +25% | 1 | 1 | 0% | 793 | 1,632 | +106% | 0 | 0 | — |
case-05 | fail→fail | 14,069 | 6,543 | -53% | 1 | 1 | 0% | 2,101 | 1,060 | -50% | 0 | 0 | — |
case-06 | fail→fail | 9,157 | 2,040 | -78% | 1 | 1 | 0% | 1,431 | 972 | -32% | 0 | 0 | — |
case-07 | fail→fail | 13,581 | 3,996 | -71% | 1 | 1 | 0% | 2,421 | 1,293 | -47% | 0 | 0 | — |
case-08 | pass→pass | 8,926 | 3,939 | -56% | 1 | 1 | 0% | 1,511 | 1,386 | -8% | 0 | 0 | — |
case-09 | fail→pass | 14,691 | 7,294 | -50% | 1 | 1 | 0% | 2,292 | 1,980 | -14% | 0 | 0 | — |
case-10 | fail→fail | 9,549 | 4,010 | -58% | 1 | 1 | 0% | 1,640 | 1,361 | -17% | 0 | 0 | — |
case-11 | fail→fail | 5,190 | 3,255 | -37% | 1 | 1 | 0% | 863 | 1,146 | +33% | 0 | 0 | — |
case-12 | fail→fail | 9,120 | 3,612 | -60% | 1 | 1 | 0% | 1,701 | 1,185 | -30% | 0 | 0 | — |
case-13 | fail→pass | 9,291 | 2,390 | -74% | 1 | 1 | 0% | 1,538 | 1,027 | -33% | 0 | 0 | — |
case-14 | fail→pass | 12,963 | 5,658 | -56% | 1 | 1 | 0% | 2,064 | 1,649 | -20% | 0 | 0 | — |
case-15 | fail→pass | 12,883 | 2,613 | -80% | 1 | 1 | 0% | 2,088 | 1,064 | -49% | 0 | 0 | — |
case-16 | fail→pass | 11,416 | 3,033 | -73% | 1 | 1 | 0% | 1,758 | 1,026 | -42% | 0 | 0 | — |
case-17 | pass→pass | 19,613 | 12,050 | -39% | 1 | 1 | 0% | 3,351 | 2,630 | -22% | 0 | 0 | — |
case-18 | pass→pass | 12,150 | 5,154 | -58% | 1 | 1 | 0% | 2,083 | 1,516 | -27% | 0 | 0 | — |
case-19 | pass→pass | 7,707 | 6,142 | -20% | 1 | 1 | 0% | 1,369 | 1,625 | +19% | 0 | 0 | — |
case-20 | fail→fail | 10,321 | 5,110 | -50% | 1 | 1 | 0% | 1,736 | 892 | -49% | 0 | 0 | — |
case-21 | fail→pass | 5,523 | 1,442 | -74% | 1 | 1 | 0% | 925 | 837 | -10% | 0 | 0 | — |
case-22 | pass→pass | 8,417 | 2,676 | -68% | 1 | 1 | 0% | 1,284 | 1,049 | -18% | 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 20 counted toward the lift figure. The other 2 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 +32 percentage points is the difference between those two pass rates over the 20 comparable cases. 3 cases got worse with the skill loaded, and they are included in that figure.
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.