CodingText outputGuided
Agent Report Card (Pass/Fail Tests)
Builds simple pass/fail tests for an AI agent and scores a run. Why it matters: if you can't test it, you can't trust it.
Scanned · no issuesReady to use
Prompt
Create an agent report card.
Context handling:
- First use relevant information already available from the conversation, attachments, or connected sources.
- Do not ask me to repeat details you already have.
- If a required detail is missing, ask one short follow-up.
- State important assumptions. Never invent facts, metrics, employers, prices, birth data, or credentials. Use [NEEDED] when something is missing.
What the agent should do:
{{task}}
Sample inputs:
{{samples}}
Output: 8–12 pass/fail checks, how to score a run, failure taxonomy, one regression case.Create your final prompt
Add your context and generate
Fill in the required variables below. Optional variables can be left blank and will remain as placeholders.
Agent job
Examples
What to expect
Example output
A clear structured answer you can act on.
Example context
Example input
Paste your details into the variables.
Coding workflow prompts
More in this series
- Create a Minimal Bug-Reproduction Plan
- Convert Requirements into Acceptance Tests
- Plan a Safe Dependency Upgrade
- Generate SQL with a Validation Plan
- Turn My Workflow Into a Cursor Skill
- 4 Agents to Spec, Build, Break, and Fix a Feature
- Generate API Test Cases from a Contract
- Review an Architecture Diagram for Risks
- REST API Design — ChatGPT Prompt
- GitHub Actions CI Pipeline — ChatGPT Prompt
- Debugging Root Cause Analysis — ChatGPT Prompt
- Hardened Dockerfile — ChatGPT Prompt
Keep exploring
Discussion
Comments (0)
Sign in to leave a comment.
Loading comments…