agentleFS
daviszhendaviszhen/agent-skills/skills/criteria-audit/SKILL.md

criteria-audit

Skill1 starsChanged 23 days agoMITView on GitHub

Audit a product, direction or idea with the three hard criteria and three axes (including the AI test), then output a filled checklist, verdict and stop-loss. Use when the user asks "is this worth doing", "should I build this", "帮我判断这个方向", or wants a go/no-go decision.

The file

We split this skill into the rules it gives an agent

7 rules. About 0 are common, 7 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈7

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    If "who pays" is unclear, say so and stop.

    line 20
  2. Distinctive:

    All three must pass; otherwise it does not produce autonomy.

    line 26
  3. Distinctive:

    Check: DIY cost ≥ 2× price.

    line 38
  4. Distinctive:

    Write time box + a single binary metric.

    line 47
  5. Distinctive:

    Never invent facts about the user's situation.

    line 66
  6. Distinctive:

    Keep the three axes separate.

    line 67
  7. Distinctive:

    If the object is content with no price, say plainly it is a channel, not a product.

    line 68

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Criteria Audit

A repeatable audit for "should I build this?" questions. The goal is not to encourage or discourage — it is to make the judgement structural.

When to use

  • User asks whether a product / direction / idea is worth doing.
  • User asks for a go/no-go decision, or a stop-loss.
  • User wants to compare two candidate directions.

Steps

  1. Compress into one line [who] pays [how much] for [what they can take and use] If "who pays" is unclear, say so and stop.
  2. Three hard criteria
    • ① Deliverable usable: can the user take something away and use it? (Content/analysis/opinion → no.)
    • ② Non-employment income: is the money paid outside an employment relationship? (Salary → no; ads/IAP ≈ 0; direct payment → yes.)
    • ③ Portable: after leaving the employer/platform, is it still theirs? All three must pass; otherwise it does not produce autonomy.
  3. Three axes (keep separate, never merge)
    • Compliance: only the three criteria above.
    • Feasibility: minimum scale to monetize, distance from current skills, free tools solving 80%, AI test.
    • Risk: single point of cut (platform/payment/compliance), AI commoditization, competition.
  4. AI test
    • Can a free LLM / code agent reach 80 points in 10 minutes? End-to-end?
    • What is the remaining 20% (accountability, context, integration, trust, compliance, maintenance) — and is any of it yours?
  5. Pricing threshold (if it charges money) Check: DIY cost ≥ 2× price; something AI can't get; verifiable; measurable; repeatable.
  6. Quadrant & decision
    • complies × feasible → do it, validate in small steps
    • complies × not feasible → change form / wait / side line
    • doesn't comply × feasible → cash flow only, not autonomy
    • doesn't comply × not feasible → drop
  7. Stop-loss Write time box + a single binary metric.

Output format

Object: ...
One line: who / pays how much / gets what
① usable: yes/no  ② non-employment: yes/no  ③ portable: yes/no
→ all three? ...

Feasibility: min scale __ / distance __ / free or AI 80%? __ / AI test: __/10
Risk: single point __ / compliance __ / AI commoditization __ / competition __

Verdict: do / cash flow only / change form / drop
Stop-loss: __ weeks, metric: __

Rules

  • Never invent facts about the user's situation; ask or mark as unknown.
  • Keep the three axes separate; do not let feasibility or risk impersonate compliance.
  • If the object is content with no price, say plainly it is a channel, not a product.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.