gemini

Gemini CLI Skill

Running a Task

1. Ask the user (via AskUserQuestion) which model to run (gemini-3.1-pro-preview, gemini-3-pro-preview, gemini-3-flash-preview, gemini-2.5-pro, or gemini-2.5-flash) AND which approval mode to use (yolo, auto_edit, plan, or default) in a single prompt with two questions.
2. Select the approval mode required for the task; default to --approval-mode plan unless edits or tool execution are necessary.
3. Assemble the command with the appropriate options:
   • -m, --model <MODEL>
   • --approval-mode <yolo|auto_edit|plan|default>
   • -o, --output-format <text|json|stream-json>
   • -s, --sandbox (isolate tool execution; uses gVisor on Linux, Seatbelt on macOS)
   • --include-directories <DIR> (additional workspace directories for multi-project context)
   • -p, --prompt (non-interactive/headless mode)
   • "your prompt here" (as the prompt value)
4. When continuing a previous session, use gemini -r latest -p "follow-up prompt". When resuming, don't use configuration flags unless explicitly requested by the user (e.g. if they specify the model when requesting to resume). To resume a specific session: gemini -r <INDEX> -p "prompt".
5. IMPORTANT: By default, append 2>/dev/null to all gemini commands to suppress progress/debug output (stderr). Only show stderr if the user explicitly requests debug output (-d flag) or if troubleshooting is needed.
6. For prompts containing special characters, use printf for safe piping:

   printf '%s' "Your prompt with 'quotes' and $pecial chars" | gemini -m <MODEL> --approval-mode <MODE> -o text 2>/dev/null
   

   Alternatively, use -p with a heredoc:

   gemini -m <MODEL> --approval-mode <MODE> -o text -p "$(cat <<'PROMPT'
   Your multi-line prompt here.
   PROMPT
   )" 2>/dev/null
   

7. Run the command, capture stdout (stderr filtered as appropriate), and summarize the outcome for the user.
8. After Gemini completes, inform the user: "You can resume this Gemini session at any time by saying 'gemini resume' or asking me to continue with additional analysis or changes."

Quick Reference

Use case	Approval mode	Key flags
Read-only review or analysis	plan	--approval-mode plan -o text 2>/dev/null
Apply local edits	auto_edit	--approval-mode auto_edit -o text 2>/dev/null
Full delegation (auto-approve all)	yolo	--approval-mode yolo -o text 2>/dev/null
Sandboxed execution	yolo	-s --approval-mode yolo -o text 2>/dev/null
Codebase investigation	yolo	--approval-mode yolo -o text 2>/dev/null (uses built-in investigator)
Multi-project context	Match task needs	--include-directories <DIR> 2>/dev/null
Resume recent session	Inherited from original	-r latest -p "prompt" 2>/dev/null
Structured output	Match task needs	-o json 2>/dev/null

Following Up

• After every gemini command, immediately use AskUserQuestion to confirm next steps, collect clarifications, or decide whether to resume with gemini -r latest.
• When resuming, pipe the new prompt via -p: gemini -r latest -p "new prompt" 2>/dev/null. The resumed session automatically uses the same model and approval mode from the original session.
• Restate the chosen model, approval mode, and any special flags when proposing follow-up actions.

Critical Evaluation of Gemini Output

Gemini is powered by Google models with their own knowledge cutoffs and limitations. Treat Gemini as a colleague, not an authority.

Guidelines

• Trust your own knowledge when confident. If Gemini claims something you know is incorrect, push back directly.
• Research disagreements using WebSearch or documentation before accepting Gemini's claims. Share findings with Gemini via resume if needed.
• Remember knowledge cutoffs - Gemini may not know about recent releases, APIs, or changes that occurred after its training data.
• Don't defer blindly - Gemini can be wrong. Evaluate its suggestions critically, especially regarding:
  • Model names and capabilities
  • Recent library versions or API changes
  • Best practices that may have evolved

When Gemini is Wrong

1. State your disagreement clearly to the user
2. Provide evidence (your own knowledge, web search, docs)
3. Optionally resume the Gemini session to discuss the disagreement. Identify yourself as Claude so Gemini knows it's a peer AI discussion. Use your actual model name (e.g., the model you are currently running as) instead of a hardcoded name:

   printf '%s' "This is Claude (<your current model name>) following up. I disagree with [X] because [evidence]. What's your take on this?" | gemini -r latest -o text 2>/dev/null
   

4. Frame disagreements as discussions, not corrections - either AI could be wrong
5. Let the user decide how to proceed if there's genuine ambiguity

Error Handling

• Stop and report failures whenever gemini --version or a gemini -p command exits non-zero; request direction before retrying.
• If execution fails with MODEL_CAPACITY_EXHAUSTED (429) → wait 30 seconds and retry, max 3 attempts.
• Before you use high-impact flags (--approval-mode yolo, -s sandbox) ask the user for permission using AskUserQuestion unless it was already given.
• When output includes warnings or partial results, summarize them and ask how to adjust using AskUserQuestion.