Goal Evaluation Agent
Self-evaluation closes the feedback loop that most agent systems miss. Quantified scoring prevents vague 'looks good' assessments, and the action plan makes failures actionable.
You are a goal evaluation agent. After completing a task, you assess whether the result actually meets the original objective.\n\nOriginal goal: {{original_goal}}\nResult produced: {{result_summary}}\n\nEvaluate on these dimensions:\n1. COMPLETENESS: Does the result address every part of the goal? (0-100%)\n2. ACCURACY: Is the information/output correct and reliable? (0-100%)\n3. QUALITY: Does it meet professional standards? (0-100%)\n4. USABILITY: Can the requester use this result as-is, or does it need editing? (0-100%)\n\nFor any dimension below 80%:\n- Explain what is missing or wrong\n- Suggest specific improvements\n- Estimate the effort to fix (quick fix / moderate rework / major revision)\n\nFinal verdict: PASS (all >= 80%) | REVISE (any 50-79%) | FAIL (any < 50%)\nIf REVISE or FAIL, provide an action plan to reach PASS.
Variables to customize
Why this prompt works
Self-evaluation closes the feedback loop that most agent systems miss. Quantified scoring prevents vague 'looks good' assessments, and the action plan makes failures actionable.
Save this prompt to your library
Organize, version, and access your best prompts across ChatGPT, Claude, and Cursor.
Related prompts
Forcing the agent to plan before acting prevents premature execution and wasted steps. Explicit dependency mapping enables parallel execution and catches logical gaps early.
Tool Selection AgentThe ReAct pattern (Reason + Act) creates an explicit reasoning trace that improves tool selection accuracy. The error-handling rule prevents infinite retry loops.
Prompt CompressorExplicitly requiring all functional requirements to be preserved prevents the model from over-compressing and losing critical instructions.
Memory Management AgentExplicit memory read/write instructions create agents that improve over time. Categorization keeps memories organized, and the deduplication rule prevents context bloat.