# Gemini

> Prompt Gemini headlessly for a second opinion, research, analysis, or code review.

- Skill: `sreejithraman/gemini` (Agent Skill, multi-file: 4 files)
- Install (CLI): `npx skillmds@latest add sreejithraman/gemini`
- Raw SKILL.md: https://api.skillmd.com/api/skills/sreejithraman/gemini/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: AI & ML
- Author: sreejithraman (https://skillmd.com/u/sreejithraman)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/sreejithraman/gemini

---


# Gemini

Give Gemini a self-contained task: the question, relevant context, scope,
constraints, and the kind of answer you need. It cannot see this conversation.
Include the facts it needs without steering it toward your preferred conclusion.
For an independent opinion, leave out earlier reviewers' conclusions.

Send the prompt headlessly through agy and wait for the answer. Use
[headless.md](references/headless.md) for headless configuration. Start fresh for a
new task; keep the same conversation when asking a follow-up.

For code review, use [review.md](references/review.md) to pass the exact scope
and optionally use Gemini's installed review skill. Keep analysis read-only
unless the caller asks for changes.

Assess the answer against the task and available evidence. Follow up when a
material gap needs clarification. Tell the caller what Gemini found, what you
recommend and why, and any limits or disagreements. An incomplete run is not
an answer.