From cli-agent-skills
This skill should be used when the user asks to "research a topic with Gemini", "use gemini-research", "delegate research to Gemini", "have Gemini look into", "ask Gemini to research", or wants a structured research report generated by Gemini CLI. Requires a topic argument.
How this skill is triggered — by the user, by Claude, or both
Slash command
/cli-agent-skills:gemini-researchThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Delegate research tasks to Gemini CLI headlessly and return structured findings without dumping raw output into Claude Code's context window.
Delegate research tasks to Gemini CLI headlessly and return structured findings without dumping raw output into Claude Code's context window.
When this skill triggers, run the bundled script with the user's topic:
"${CLAUDE_PLUGIN_ROOT}/skills/gemini-research/scripts/gemini-research.sh" "<topic>"
Replace <topic> with the subject the user wants researched. Quote the topic argument — it may contain spaces.
gemini -p "<prompt>" headlessly~/.claude/cli-reviews/Read the markdown printed to stdout and present it to the user. Do not include the raw script invocation in the response — only the formatted findings.
If the script exits non-zero, report the error message it printed. Do not attempt to fall back to your own research unless the user explicitly asks.
Each run writes one JSON file to ~/.claude/cli-reviews/gemini-research-<topic-slug>-<timestamp>.json. Mention the artifact path at the end of your response so the user knows where findings are saved.
Creates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.
npx claudepluginhub mrjamescollins/claude-plugins --plugin cli-agent-skills