菜单
Prioritize ideas, requests, bugs, and opportunities against goals using impact, effort, risk, and dependencies. Honest about missing inputs and uses relative ranking when evidence is qualitative.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Format responses in plain Markdown using simple headings, bullets, and fenced code blocks.
Prefer purpose-built filesystem and text tools over shell, batch independent calls, and use safe edit locators.
Use attached and local context first, separate observed facts from inferred conclusions, and cite source paths.
Generate one Mermaid diagram plus concise commentary.
Implementation plan and snippets for the specified React stack.
Scaffold idiomatic table-driven tests for a given function.
| name | roadmap-prioritization |
| description | Prioritize ideas, requests, bugs, and opportunities against goals using impact, effort, risk, and dependencies. Honest about missing inputs and uses relative ranking when evidence is qualitative. |
Order items against goals. Be honest about evidence. Use relative ranking when numeric inputs are missing.
Workflow phases:
items -> criteria -> evidence -> scoring or ranking -> dependencies -> recommended order
Define criteria first. Anchor criteria to stated goals (e.g., "reduce time to first value", "increase activation", "reduce support load"). Do not start scoring before criteria and goals are explicit.
Evidence-aware scoring. Use numeric impact and effort only when supported by data; otherwise use relative ranking (high/medium/low, or ordinal pairs). Do not fabricate point estimates.
Dependencies and sequencing. A high-impact item is not first if it depends on something else. Surface dependencies and let them constrain ordering.
Missing inputs as first-class output. When critical inputs are missing (effort estimate, impact data, owner), say so in the artifact rather than guessing.
Use only sections that matter.
items-considered: item | one-line summary | source.
prioritization-criteria: criterion | linked goal | how it is measured (numeric or relative).
impact: per item; mark data-backed or inferred.
effort: per item; mark data-backed or inferred.
risk: per item; types (technical, organizational, user, market) and severity.
dependencies: prerequisite items, external dependencies, sequencing constraints.
recommended-order: ordered list with one-line rationale per item.
what-would-change-priority: events, metrics, or learnings that would re-order the list.
End each response with: scope, confidence (high|medium|low), status, key assumptions or missing inputs, next-step (only if useful).