AI Agent System Prompt
Build a system prompt for an autonomous, tool-using AI agent — plan before acting, one tool at a time, act on real results, stop before irreversible steps.
View Resource →Prompt Builders
System Prompt Generator
Configure role, behavior rules, and output structure for your AI system. Generates a structured system prompt ready to paste into any AI model. Runs entirely in your browser.
Model-agnostic structure that works across any instruction-following model.
Behaviors the AI must consistently follow. One rule per line.
Hard limits and out-of-scope behaviors. One rule per line.
When the AI should express uncertainty or defer to a human.
Show the AI what good input and expected output look like.
Response Style
Word limit adds an instruction for future AI replies — it does not shorten this generated prompt. Choose a value between 50 and 5000 words.
Generated Prompt
Try with:
Bug Triage Assistant
Convert scattered bug notes, Slack messages, or user complaints into structured engineering tasks with reproduction steps, severity, and root cause hypothesis.
View Resource →Code Review Assistant
Turn AI into a structured pull request reviewer that catches risky changes, flags maintainability issues, and suggests missing test coverage.
View Resource →Customer Support Agent
Configure AI to answer support questions within your actual policy boundaries — not generic best-guess answers.
View Resource →Escalation Response Assistant
Structure the escalation decision for a high-risk support situation — consistent criteria, clear rationale, and a ready brief for the team taking over.
View Resource →Internal Knowledge Base Assistant
Answer questions using your KB content directly — citing the source, staying within documented scope, and refusing to fill gaps with guesses.
View Resource →Meeting Notes Structurer
Extract decisions, action items, and open questions from raw meeting notes — the three things that need to survive the meeting.
View Resource →Product Requirements Assistant
Turn a rough feature idea into a structured requirement: problem statement, acceptance criteria, and what's explicitly out of scope.
View Resource →Research Synthesis Assistant
Extract and compare findings from multiple sources without collapsing them into a single blended perspective.
View Resource →SEO Content Brief Assistant
Build a structured content brief from a keyword or topic before drafting — so writers start with intent, not assumptions.
View Resource →Technical Summary Assistant
Compress long technical content into summaries calibrated to a specific audience and purpose — not a generic restatement.
View Resource →Test Case Generator
Generate edge-case-focused test scenarios for any function, feature, or API endpoint — covering boundary values, failure modes, and regression risks.
View Resource →AI System Prompt Design Workflow
Design a system prompt that holds up in production — define the role precisely, engineer the behavior and guardrails on top of it, then check it reads clearly before you ship.
View Playbook →AI Agent Instruction Workflow
Instruct an AI agent that runs on its own without it wandering off — anchor it to a role, write the agent system prompt, then lay out the multi-step plan it works through.
View Playbook →AI Research Synthesis Workflow
Pull a single coherent view out of a stack of sources — package them together, summarize each faithfully, then have AI synthesize across them instead of one at a time.
View Playbook →AI Customer Support Workflow
Run inbound support the same way every time — triage and route the ticket, pull the details that matter, draft a reply in a consistent voice, and log the resolution for the record.
View Playbook →How it works
Fill in the role and objective, describe the operational context, and choose tone and output format. The Advanced section lets you specify must-do rules, hard restrictions, escalation handling, and example inputs/outputs. Click Generate Prompt to produce a structured system prompt from your inputs. Everything runs in your browser — nothing is sent to a server.
Use cases
- Configuring consistent AI behavior for engineering review workflows
- Building support agent prompts that stay within documented policy boundaries
- Defining research synthesis instructions that preserve source attribution
- Generating starting-point prompts for internal tools, prototypes, or team use
Pro tips
- Role and Objective work together — the role defines what the AI is, the objective defines what it's trying to accomplish in this specific context.
- The Must Not Do section is often more valuable than Must Do. Explicit restrictions prevent the AI from improvising in ways you haven't anticipated.
- Use the Example Inputs/Outputs field to anchor abstract instructions. A concrete example does more to shape output quality than three extra rules.
- For Claude, the XML Tags format reduces hallucination on structured tasks. For ChatGPT, Markdown Sections tends to produce clean output.
FAQ
Does this tool send my inputs to an AI model?
No. All prompt generation happens in your browser using JavaScript. Nothing is sent to a server or external API.
What is the difference between Must Do and Rules?
They're the same thing — Must Do instructions become the Rules section in your generated prompt. The field name describes the intent: behaviors your AI system must consistently follow.
Which models work with the generated prompts?
The Target Model selector formats output for Claude, ChatGPT/OpenAI, Gemini, or a generic format that works with any standard instruction-following model. The content is model-agnostic — the format differs slightly.
Can I edit the generated prompt after it's produced?
Yes. The output area is fully editable. Adjust the generated prompt directly before copying or downloading.
What does the token estimate mean?
The token count is an approximation using a 1 token ≈ 4 characters heuristic. Actual counts vary by model and tokenizer — treat it as a rough guide, not a precise measurement.
What does Prompt Depth mean?
Depth is a heuristic that reflects how many sections of the form are filled. Basic means only role is defined. Solid means role, context, and some rules. Detailed means role, context, rules, restrictions, and additional guidance are all populated.