CtrlK
BlogDocsLog inGet started
Tessl Logo

gemini

Interface with the Gemini agent for operations requiring the Gemini model.

Invalid
This skill can't be scored yet
Validation errors are blocking scoring. Review and fix them to unlock Quality, Impact and Security scores. See what needs fixing →
SKILL.md
Quality
Evals
Security

Gemini Agent

Interact with the Gemini agent. This skill wraps input in strict JSON enforcement prompts and handles retries.

Routing

Image generation? Use the gemini-image-generator skill instead. This skill is for text/reasoning tasks only.

Activation

Triggers: "ask gemini", "gemini help", "consult gemini"

Features

  • Strict JSON Output: The agent is prompted to return ONLY valid JSON.
  • Retry Logic: Automatically retries with stricter prompts if the first response is not valid JSON.
  • Schema Validation: Supports passing a JSON schema to validate the output against (if supported by the underlying implementation).

Usage

Simple Prompt

# Usage: .claude/skills/gemini/gemini-agent.sh "Your prompt here"
.claude/skills/gemini/gemini-agent.sh "Summarize the history of AI."

JSON Request (Advanced)

You can pass a JSON object with prompt, context, and schema.

{
  "prompt": "Generate a user profile",
  "schema": {
    "type": "object",
    "properties": {
      "username": {"type": "string"},
      "age": {"type": "integer"}
    }
  }
}
.claude/skills/gemini/gemini-agent.sh '{"prompt": "...", "schema": {...}}'

Response Format

Success:

{
  "key": "value",
  ...
}

Error:

{
  "ok": false,
  "error": {
    "code": "error_code",
    "message": "Description"
  }
}
Repository
stephenfeather/claude-skills
Last updated
First committed

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.