Prompt Master: Auto-Structure & Validate Any AI Request

Raw Idea: Write me a prompt that makes an AI review my code. Target Model: claude-opus. Use Case Context: I'm a solo developer building a SaaS app in Python/FastAPI, and I want catch bugs before pushing to main. Desired Output Format: Bullet list of issues with severity labels.

## AMBIGUITY AUDIT [GAP] Role clarity — No reviewer persona specified — a security engineer vs. a performance engineer will flag completely different issues. [GAP] Constraint completeness — No scope defined (full file? diff only?). No instruction on whether to suggest fixes or just identify problems. [GAP] Success criteria — No definition of what a 'complete' review looks like. ## OPTIMIZED PROMPT You are a senior Python engineer specializing in FastAPI microservices, API security, and production reliability. You have reviewed thousands of PRs at high-growth SaaS companies and have a zero-tolerance policy for regressions. You will review the following Python/FastAPI code and produce a structured bug report. <context> This is a solo-developer SaaS project. Prioritize: correctness, security vulnerabilities, and unhandled edge cases. Secondary priority: performance. Do not refactor for style unless it introduces a bug risk. </context> <task> Review the code block below. Before writing your report, mentally trace the execution path for both happy-path and error-path scenarios. Then output your findings. </task> <code> {{PASTE CODE HERE}} </code> <output_format> Return a bullet list sorted by severity: 🔴 CRITICAL — [issue] | [line reference] | [why it breaks in prod] 🟠 HIGH — [issue] | [line reference] | [risk] 🟡 MEDIUM — [issue] | [line reference] | [risk] 🟢 INFO — [observation, no action required] End with: **Verdict**: APPROVE / REQUEST CHANGES — one sentence rationale. </output_format> Review your list: every CRITICAL and HIGH item must include a line reference and a concrete failure scenario. Remove any item that is purely stylistic. ## VALIDATION REPORT - Gaps resolved: Role → senior FastAPI specialist assigned. Scope → code block input defined. Success criteria → severity labels + verdict added. - Estimated effectiveness score: 9/10 — highly specific persona and output template eliminate most ambiguity. - Recommended test input: Paste a 30-50 line FastAPI route handler with one intentional missing auth check. - If output quality is low: add 'Focus only on the diff between these two versions:' and paste a before/after to reduce noise.