Gopherlinzy

PRD-to-code conflict detection. Activate when: (1) user asks to audit/check/validate a PRD against existing code, (2) user says "prd:audit", "审 PRD", "需求冲突检测", "PRD 审查", (3) user provides a PRD document and asks if it conflicts with current implementation, (4) before Spec Coding to validate requirements feasibility. Output: structured audit report with CONFLICT/WARNING/GAP/PASS findings, each backed by code evidence (file path + line numbers). NOT for: plan generation, code review, writing PRDs, or tasks better handled by spec-kit / Superpowers.

对文章内容、技术方案、用户提问进行批判性红队审查，找出时效问题、逻辑漏洞、错误假设和幻觉引用。触发场景：用户说「红队」「挑战我的方案」「反驳一下」「质疑一下」「帮我找问题」「我说的有没有问题」「这个方案靠谱吗」「有没有更好的方案」，或提交文章/方案/技术设计请求审查时。支持三种模式：A内容审查（文章可靠性）、B方案审查（逻辑漏洞/安全问题）、C前提审查（挑战错误假设/认知偏差）。

PRD-to-code conflict detection. Activate when: (1) user asks to audit/check/validate a PRD against existing code, (2) user says "prd:audit", "审 PRD", "需求冲突检测", "PRD 审查", (3) user provides a PRD document and asks if it conflicts with current implementation, (4) before Spec Coding to validate requirements feasibility. Output: structured audit report with CONFLICT/WARNING/GAP/PASS findings, each backed by code evidence (file path + line numbers). NOT for: plan generation, code review, writing PRDs, or tasks better handled by spec-kit / Superpowers.