gemini-prompting

Gemini Prompting

Use this skill when gemini:gemini-rescue needs to ask Gemini for help.

Prompt Gemini like an operator, not a collaborator. Keep prompts compact and block-structured with XML tags. State the task, the output contract, the follow-through defaults, and the small set of extra constraints that matter.

Core rules:

• Prefer one clear task per Gemini run. Split unrelated asks into separate runs.
• Tell Gemini what done looks like. Do not assume it will infer the desired end state.
• Add explicit grounding and verification rules for any task where unsupported guesses would hurt quality.
• Prefer better prompt contracts over long natural-language explanations.
• Use XML tags consistently so the prompt has stable internal structure.

Default prompt recipe:

• <task>: the concrete job and the relevant repository or failure context.
• <output_contract>: exact shape, ordering, and brevity requirements.
• <follow_through_policy>: what Gemini should do by default instead of asking routine questions.
• <verification_loop>: required for debugging, implementation, or risky fixes.
• <grounding_rules>: required for review, research, or anything that could drift into unsupported claims.

When to add blocks:

• Coding or debugging: add verification_loop and missing_context_gating.
• Review or adversarial review: add grounding_rules, output_contract, and emphasis on material findings only.
• Research or recommendation tasks: add research_mode and citation requirements.
• Write-capable tasks: add action_safety so Gemini stays narrow and avoids unrelated refactors.

Leverage Gemini's strengths:

• Large context window (1M tokens) — include full file contents when relevant, not just snippets.
• Strong reasoning — ask for explicit step-by-step analysis on complex architectural or security questions.
• Multimodal — if the task involves diagrams, screenshots, or visual artifacts, include them.

Working rules:

• Prefer explicit prompt contracts over vague nudges.
• Use stable XML tag names.
• Ask Gemini for brief, outcome-based progress updates only when the task is long-running.
• Keep claims anchored to observed evidence. If something is a hypothesis, say so.

How to choose prompt shape:

• Use built-in review or adversarial-review commands when the job is reviewing local git changes. Those prompts already carry the review contract.
• Use task when the task is diagnosis, planning, research, or implementation and you need to control the prompt more directly.
• Use task --resume-last for follow-up instructions on the same Gemini thread. Send only the delta instruction instead of restating the whole prompt unless the direction changed materially.

Prompt assembly checklist:

1. Define the exact task and scope in <task>.
2. Choose the smallest output contract that still makes the answer easy to use.
3. Decide whether Gemini should keep going by default or stop for missing high-risk details.
4. Add verification, grounding, and safety tags only where the task needs them.
5. Remove redundant instructions before sending the prompt.

Reusable blocks live in references/prompt-blocks.md. Concrete end-to-end templates live in references/gemini-prompt-recipes.md. Common failure modes to avoid live in references/gemini-prompt-antipatterns.md.