Generic Output Debugger

Fix prompts that produce generic or unhelpful AI responses

analysisbeginnerCRISP FrameworkPrompt Engineering400-500 words

Customize Your Prompt

Fill in the variables to generate your personalized prompt

Your Original Prompt

What You're Trying to Achieve

Additional Context (Optional)

Preview

See how your prompt will look with the current variables

Your prompt "[Your Original Prompt]" is producing generic responses for [What You're Trying to Achieve].

Role: Expert Prompt Engineer specializing in CRISP methodology and PM workflows.

Instructions:
1. Analyze prompt for CRISP framework gaps
2. Identify missing context and vague instructions
3. Rewrite using CRISP structure
4. Add constraints and examples
5. Include success criteria

Specifics:
## Diagnosis
**Missing Elements:**
- **Context:** [What background is missing]
- **Role:** [How role could be more specific]
- **Instructions:** [Where steps are unclear]
- **Specifics:** [What format is undefined]
- **Purpose:** [How purpose could guide better]

## Improved CRISP Prompt
**Context:** [Enhanced context]
**Role:** [Specific expertise]
**Instructions:** [Clear numbered steps]
**Specifics:** [Exact format and constraints]
**Purpose:** [Clear use case]

## Why This Works Better
- [Specific improvement 1]
- [Specific improvement 2]
- [Specific improvement 3]

## Test Validation
**Good Output Would Include:** [Specific elements]
**Red Flags to Avoid:** [What indicates failure]

Purpose: Transform vague prompts into precise, actionable ones.

## Evidence and accuracy
- Use supplied facts and verified sources. Do not invent metrics, quotes, people, company details, commitments, or personal experience.
- Mark assumptions [ASSUMPTION], estimates [ESTIMATE: method], and unresolved questions [UNCERTAIN: reason]. Leave unavailable values unfilled rather than guessing.
- Explain confidence as high, medium, or low using the evidence available. These labels are judgments, not calibrated probabilities. Give numerical probabilities only when a stated method supports them.
- Treat preset weights, scores, timelines, and targets in this template as starting examples. Adapt them to the task and explain changes; they are not universal benchmarks or approved commitments.
- Define score scales and directions before calculating totals. Keep units, denominators, and time periods consistent. Do not average away a critical blocker.
- Use only exact supplied or verified quotes with attribution. Label requested fictional examples as illustrative. Distinguish observed behavior from inferred motives or causes.
- Use only the sections and rows the task needs. Write plainly, preserve necessary technical terms, and avoid unsupported benefits or forced specificity.
- Identify missing information needed for a decision. Recommendations remain proposals until reviewed by the responsible team.

What to check in a prompt

• State the context, role, instructions, specifics, and purpose (CRISP).
• Include constraints and examples of the output you need.
• Define success criteria and errors to check in the output.
• Make one request per prompt, or separate requests into clear sections.
• Test small changes to see whether they improve the output.

Common Prompt Debugging Mistakes

• Adding length without clarifying the request or constraints.
• Combining writing, analysis, and decisions without separating the tasks.
• Omitting examples of the required output format.
• Leaving out the purpose, so relevant content may still miss the task.
• Skipping edge cases when testing the prompt.

Prompt debugging questions

How can I make the answers more useful?

Check whether the prompt supplies the context, constraints, examples, and success criteria the task needs. Add relevant facts you have; leave missing facts out rather than asking the model to invent them.

Can I ask for multiple things at once?

Yes. Give each request a section and an output format, or split the work into separate prompts.

How do I know if the output is “good”?

List what the output must include and the mistakes to avoid. Check the result against those requirements, then revise the prompt where it was unclear.

How to use this prompt