AI Opportunity Evaluation
A go/no-go gate for TranFu product ideas before PRD or development. Runs a 2-3 round short-question triage on a one-liner idea and outputs a one-screen decision card (Go / Refine first / Don't do).
Basic info
- Name
- AI Opportunity Evaluation
- Description
- 当用户说"判断这个项目怎么样" / "立项判断" / "这个方向能不能做" / "这个产品想法值不值得做" / "用 AI机会立项判断 Skill,帮我分析"时,做 TranFu 产品想法立项会诊,短问后输出一屏以内的立项判断卡。Do NOT trigger when 用户已经明确要求直接开发、修 bug、部署、写代码,或已完成立项并要求写 PRD。
AI Opportunity Evaluation
Triages a one-liner product idea before PRD or development. Asks 2-3 short questions, then outputs a one-screen decision card so a raw idea does not slide straight into a low-quality demo.
When to use it
- A teammate has only a product name or a one-liner idea and needs a quick worth-doing verdict.
- An internal tool is about to enter PRD or development, and the real problem and validation goal must be confirmed first.
- You want to judge whether a project can be distilled into a TranFu skill, agent, SOP, website product, or external case study.
Sibling skill comparison
Drafted by tranfu-publish, signed off by author / recommender. Helps a reader pick across siblings or jump to a better fit.
Inside the company library
- project-scoring — Quantitative scoring and decision memo for AI-workflow projects; this skill differs: it is a one-screen upstream gate, not a full scoring report.
- ai-startup-feasibility-check — Feasibility self-check for AI startup directions; this skill differs: it focuses on TranFu internal projects, capability validation, and brand risk.
- strategy-first-development — Locks strategy, product shape, and tech plan before development; this skill differs: it runs even earlier — deciding whether the project should reach PRD at all.
Outside the company
- None known.
Unique value of this skill
- Phase-two go/no-go gate.
- One-screen decision card.
- Explicit check on TranFu capability precipitation.
Usage tips
Drafted through tranfu-publish (author / recommender answers, AI integrates, recommender signs off). Helps readers ramp up vertically — tacit knowledge lives here. See §Sibling skill comparison above for horizontal choice.
What to feed in
- Start with the project name and a one-liner idea.
- Add who will use it and how it is solved today.
- State the validation flow, skill, or SOP you want to prove.
Recommended usage
- Evaluate one project idea per session.
- Accept 2-3 short follow-up questions.
- Only enter PRD after the verdict is Go.
Known limits
- Not a substitute for commercial due diligence or investment review.
- Does not directly produce a PRD or development plan.
- Confidence is low when input information is thin.
What you will get
- A definite verdict: Go / Refine first / Don't do.
- A short card: core reasons, real problem, validation goal, v1 scope, risk flags, and next step.
Version info
Local public Skill catalog snapshot, showing only public-safe fields.
Skill files
(6)SKILL.md
SKILL.md · Markdown