gemini
Gemini Skill
Priorities
Approval mode correctness > Model selection > Background execution safety
Model Registry
Load current models before executing — this overrides any model names in the tables below:
Glob(pattern: "**/sdlc/**/config/model-registry.md", path: "~/.claude/plugins")→ Read result- Use
gemini-flagshipas the default model. Offer usergemini-fastfor speed-critical tasks. - If registry load fails, fall back to the tables below.
Goal
Execute Gemini CLI for comprehensive code review, plan analysis, or large-context processing tasks. Ask user for model selection via AskUserQuestion. Choose approval mode based on execution context (yolo for background, default for interactive terminal only). Load CLI reference for detailed command patterns, troubleshooting, and use cases.
Constraints
- NEVER use
--approval-mode defaultin background or non-interactive shells (Claude Code tool calls). It hangs indefinitely waiting for user input. - ALWAYS use
--approval-mode yolofor automated/background tasks or wrap with timeout:timeout 300 gemini ... - Ask user for model selection via AskUserQuestion before running commands.
- After Gemini completes, inform user they can start a new session for follow-up analysis.
Model Selection (fallback — prefer registry)
Ask user which model to use via AskUserQuestion:
| Model | Best for | Context window | Key features |
|---|---|---|---|
gemini-3-pro-preview |
Flagship: Complex reasoning, coding, agentic tasks | 1M input / 64k output | Vibe coding, 76.2% SWE-bench, $2-4/M input |
gemini-3-flash |
Sub-second latency, speed-critical applications | 1M input / 64k output | Distilled from 3 Pro, TPU-optimized |
gemini-2.5-pro |
Legacy: Strong all-around performance | 1M input / 65k output | Thinking mode, mature stability |
gemini-2.5-flash |
Legacy: Cost-efficient, high-volume tasks | 1M input / 65k output | Best price ($0.15/M), thinking mode |
gemini-2.5-flash-lite |
Legacy: Fastest processing, high throughput | 1M input / 65k output | Maximum speed, minimal latency |
These model names may be outdated. Always prefer model-registry.md values when available.
References
Load CLI reference for detailed command patterns, troubleshooting hung processes, and common use cases:
Glob(pattern: "**/sdlc/**/skills/gemini/references/gemini-cli-reference.md", path: "~/.claude/plugins")→ Read result
Arguments
$ARGUMENTS
More from iamladi/cautious-computing-machine--sdlc-plugin
codex
Use when the user asks to run Codex CLI (codex exec, codex resume) or references OpenAI Codex for code analysis, refactoring, or automated editing. Resolves the latest flagship model from the model registry.
9interview
Interview me about anything in depth
7tdd
TDD enforcement during implementation. Reads `tdd:` setting from CLAUDE.md. Modes - strict (human approval for escape), soft (warnings), off (disabled). Auto-invoked by /implement.
6x-search
Search X/Twitter for real-time developer discourse, product feedback, community sentiment, and expert opinions. Use when user says "x search", "search x for", "search twitter for", "what are people saying about", or needs recent X discourse for context (library releases, API changes, product launches, industry discussion). Also use when researching a library, framework, API, or product to supplement web search with real-time community signal — e.g. "research Bun", "what do devs think of Hono", "is Turso production-ready".
1judgment-eval
Evaluates agent judgment quality through scenario-based testing in-conversation. Use when the user wants to test, validate, or stress-test an agent, skill, or command definition — e.g. "test this agent", "evaluate this skill", "does this prompt handle edge cases", "check this agent's judgment", or after writing or modifying any agent/skill/command .md file.
1update-models
Re-resolve the model registry by querying OpenAI Codex cache, Google AI API, and Oracle CLI. Use when models feel stale or after a major model release.
1