Use this shape when drafting or refactoring prompts:
Role: [who the agent is]
Goal: [single primary outcome]
Inputs: [what is available]
Constraints: [must-follow rules]
Process: [ordered steps]
Output: [exact format and fields]
Verification: [how to self-check before final]
Fallback: [what to do if blocked]- Draft the smallest viable prompt.
- Test on one typical case and one edge case.
- Identify failure mode (ambiguity, missing constraints, wrong format).
- Patch only the failing section.
- Re-test and keep a short changelog of improvements.
- Over-broad goal: multiple objectives with no priority.
- Soft constraints: words like "try" where strict behavior is needed.
- Hidden assumptions: expected context not provided in inputs.
- Output drift: no schema or quality gate before final answer.
- Convert vague goals into one measurable success condition.
- Replace optional language with explicit must/should rules.
- Add a fixed output template for structured responses.
- Include a missing-information policy (ask, infer, or stop).