Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Execute Gemini CLI for AI-powered code analysis and generation. Use when you need to leverage Google's Gemini models for complex reasoning tasks.
.claude/skills/microck-gemini/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 202% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 59% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -25% | 0% |
| case-10 | ✗→✓ | ▲ Improved | -24% | 0% |
| case-11 | ✗→✓ | ▲ Improved | 28% | 0% |
Execute Gemini CLI commands with support for multiple models and flexible prompt input. Integrates Google's Gemini AI models into Claude Code workflows.
Mandatory: Run via uv with fixed timeout 7200000ms (foreground):
bashuv run ~/.claude/skills/gemini/scripts/gemini.py "<prompt>" [working_dir]
Optional (direct execution or using Python):
bash~/.claude/skills/gemini/scripts/gemini.py "<prompt>" [working_dir] # or python3 ~/.claude/skills/gemini/scripts/gemini.py "<prompt>" [working_dir]
gemini-3-pro-preview)export GEMINI_MODEL=gemini-3timeout: 7200000 for double protectionprompt (required): Task prompt or questionworking_dir (optional): Working directory (default: current directory)Plain text output from Gemini:
textModel response text here...
Error format (stderr):
textERROR: Error message
When calling via Bash tool, always include the timeout parameter:
yamlBash tool parameters: - command: uv run ~/.claude/skills/gemini/scripts/gemini.py "<prompt>" - timeout: 7200000 - description: <brief description of the task>
Alternatives:
yaml# Direct execution (simplest) - command: ~/.claude/skills/gemini/scripts/gemini.py "<prompt>" # Using python3 - command: python3 ~/.claude/skills/gemini/scripts/gemini.py "<prompt>"
Basic query:
bashuv run ~/.claude/skills/gemini/scripts/gemini.py "explain quantum computing" # timeout: 7200000
Code analysis:
bashuv run ~/.claude/skills/gemini/scripts/gemini.py "review this code for security issues: $(cat app.py)" # timeout: 7200000
With specific working directory:
bashuv run ~/.claude/skills/gemini/scripts/gemini.py "analyze project structure" "/path/to/project" # timeout: 7200000
Using python3 directly (alternative):
bashpython3 ~/.claude/skills/gemini/scripts/gemini.py "your prompt here"
uv run for automatic Python environment management (requires uv installed)./gemini.py (uses system Python via shebang)GEMINI_MODEL environment variable)| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 8,499 | 37,676 | +343% | 1 | 1 | 0% | 862 | 2,603 | +202% | 0 | 0 | — |
case-02 | fail→fail | 8,570 | 7,790 | -9% | 1 | 1 | 0% | 1,392 | 1,346 | -3% | 0 | 0 | — |
case-03 | fail→fail | 14,429 | 6,405 | -56% | 1 | 1 | 0% | 2,437 | 1,173 | -52% | 0 | 0 | — |
case-04 | fail→pass | 4,815 | 2,939 | -39% | 1 | 1 | 0% | 810 | 1,289 | +59% | 0 | 0 | — |
case-05 | pass→pass | 4,705 | 2,463 | -48% | 1 | 1 | 0% | 787 | 1,259 | +60% | 0 | 0 | — |
case-06 | pass→pass | 3,981 | 1,506 | -62% | 1 | 1 | 0% | 604 | 1,057 | +75% | 0 | 0 | — |
case-07 | pass→pass | 6,488 | 3,148 | -51% | 1 | 1 | 0% | 1,051 | 1,412 | +34% | 0 | 0 | — |
case-08 | pass→pass | 3,924 | 1,813 | -54% | 1 | 1 | 0% | 600 | 1,013 | +69% | 0 | 0 | — |
case-09 | fail→pass | 7,907 | 1,564 | -80% | 1 | 1 | 0% | 1,317 | 988 | -25% | 0 | 0 | — |
case-10 | fail→pass | 8,510 | 1,396 | -84% | 1 | 1 | 0% | 1,321 | 1,000 | -24% | 0 | 0 | — |
case-11 | fail→pass | 5,367 | 1,992 | -63% | 1 | 1 | 0% | 882 | 1,126 | +28% | 0 | 0 | — |
case-12 | pass→pass | 7,544 | 3,973 | -47% | 1 | 1 | 0% | 1,268 | 1,561 | +23% | 0 | 0 | — |
case-13 | fail→pass | 8,392 | 3,102 | -63% | 1 | 1 | 0% | 1,480 | 1,293 | -13% | 0 | 0 | — |
case-14 | pass→pass | 4,045 | 1,586 | -61% | 1 | 1 | 0% | 562 | 1,025 | +82% | 0 | 0 | — |
case-15 | pass→pass | 5,776 | 2,096 | -64% | 1 | 1 | 0% | 883 | 1,094 | +24% | 0 | 0 | — |
case-16 | fail→pass | 10,213 | 5,323 | -48% | 1 | 1 | 0% | 1,606 | 1,589 | -1% | 0 | 0 | — |
case-17 | pass→pass | 4,204 | 3,677 | -13% | 1 | 1 | 0% | 636 | 1,400 | +120% | 0 | 0 | — |
case-18 | pass→pass | 6,402 | 2,085 | -67% | 1 | 1 | 0% | 992 | 1,153 | +16% | 0 | 0 | — |
case-19 | pass→pass | 6,315 | 2,908 | -54% | 1 | 1 | 0% | 1,103 | 1,357 | +23% | 0 | 0 | — |
case-20 | pass→pass | 7,469 | 6,215 | -17% | 1 | 1 | 0% | 1,511 | 2,025 | +34% | 0 | 0 | — |
case-21 | pass→pass | 3,703 | 2,844 | -23% | 1 | 1 | 0% | 611 | 1,223 | +100% | 0 | 0 | — |
case-22 | pass→pass | 6,163 | 5,374 | -13% | 1 | 1 | 0% | 1,082 | 1,824 | +69% | 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. 1 case got worse with the skill loaded, and it is 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.