Home › Prompts › Grok › Grok tool-using research agent prompt

Grok tool-using research agent prompt

A system prompt for a Grok agent that plans, searches, verifies and writes a cited brief, with a cap on tool calls.

GrokAgents & automationRuleStack original

Written for Grok 4.7 · Updated 2026-10-16

Note: Review and test before use. Provided as is, without warranty. See Terms.

Prompt

Download  
You are a research agent. Goal: [question]. Work in a loop: plan the sub-questions, search web and X, read primary sources, verify each key claim with at least two independent sources, then write the brief. Limit: [n] searches. Output: summary (5 bullets), detailed findings with citations, conflicting evidence, confidence (high/medium/low) per finding, and open questions.

Replace everything in [square brackets] with your own details.

Settings & parameters

Variations

Tips

More: all Grok prompts · all agents & automation prompts · prompt library

Related prompts

ChatGPT agent: multi-step web task with checkpoints

A brief for ChatGPT's agent to research, compare and compile results on the web, pausing before any purchase or submission.

ChatGPTAgents & automation

Agentic coding task brief for Claude Code

A task brief for Claude Code that defines the goal, constraints, verification and stopping point so the agent works autonomously but safely.

ClaudeAgents & automation

Gemini automation across Gmail, Docs and Calendar

A request that has Gemini gather info from email and docs and produce a draft plan, without sending anything.

GeminiAgents & automation

Design an automation workflow (n8n/Zapier/Make)

Turn a manual process into a step-by-step automation design with triggers, nodes, error handling and an LLM step prompt.

Any chat AIAgents & automation

Get new rules weekly

One email a week: fresh rule files, agent tips, and what's changing in Cursor, Claude Code and Copilot. Free, unsubscribe anytime.

Opens the Rulestack Newsletter on beehiiv in a new tab. Or subscribe directly on beehiiv.