跳至正文
OpenAgentSkill
浏览已开放 · 目录记录日期 2026-11-21 · 自动更新和账户服务暂时关闭。
设计与 UI

deterministic-metric-design

Inventing deterministic metrics — turning a fuzzy property like 'maintainability', 'risk', or 'how reducible this code is' into a deterministic, computable number an agent can trust and optimize. Covers the path from construct to adoption — operationalizing the construct, confronting computability limits (Kolmogorov, Rice) with sound proxies, picking the right measurement scale, proving properties (monotonicity, invariance, the Weyuker/Briand axioms), guaranteeing determinism, establishing construct validity (not just LOC in disguise), and hardening against Goodhart-style gaming when an agent optimizes the metric. Trigger when designing, reviewing, or validating a quantitative metric, score, measure, or index — and even when the user doesn't say 'metric' but wants to quantify, score, rank, or measure code/behavior, build a deterministic optimization target, or invent a measure for something previously unquantified (e.g., behavior-preserving codebase-size reduction).

pproenca · 已有 AI 审核记录

安装与使用

运行命令前先阅读来源与依赖说明。复制命令不会自动安装或执行。

npx skills add pproenca/dot-skills --skill deterministic-metric-design

技能说明


name: deterministic-metric-design description: Inventing deterministic metrics — turning a fuzzy property like 'maintainability', 'risk', or 'how reducible this code is' into a deterministic, computable number an agent can trust and optimize. Covers the path from construct to adoption — operationalizing the construct, confronting computability limits (Kolmogorov, Rice) with sound proxies, picking the right measurement scale, proving properties (monotonicity, invariance, the Weyuker/Briand axioms), guaranteeing determinism, establishing construct validity (not just LOC in disguise), and hardening against Goodhart-style gaming when an agent optimizes the metric. Trigger when designing, reviewing, or validating a quantitative metric, score, measure, or index — and even when the user doesn't say 'metric' but wants to quantify, score, rank, or measure code/behavior, build a deterministic optimization target, or invent a measure for something previously unquantified (e.g., behavior-preserving codebase-size reduction).

dot-skills Deterministic Metric Design Best Practices

Design metrics that are deterministic, computable, provable, and valid — measures an agent can trust and optimize against without gaming them. The 44 rules across 8 categories take a metric from a fuzzy construct to an adoptable, machine-checkable number: define the construct, confront computability limits with sound proxies, ground it in measurement theory, prove its properties, pin its determinism, validate it empirically, harden it against optimization pressure, and package it for adoption.

A running example threads through every category — a deterministic measure of behavior-preserving codebase-size reduction (shrink code without changing how the app works). It is the ideal stress test because its ideal form is provably out of reach (Kolmogorov complexity is uncomputable; program equivalence is undecidable by Rice's theorem), so the whole craft is building a deterministic, tractable proxy with a proven guarantee.

This is the measurement-design layer that the *-algorithms skills apply (Big-O, NDCG, cyclomatic, MoJoFM) but never teach.

When to Apply

Use this skill when:

  • Designing a new metric, score, or index — or reviewing someone's proposed metric for rigor
  • Asked to "quantify", "measure", "score", or "rank" a property that has no agreed measure yet
  • Building a deterministic optimization target an agent will push on (e.g., reduce code size without changing behavior)
  • Auditing an existing metric that "feels off" — it suspiciously tracks LOC, jumps between runs, or gets gamed
  • Turning a research idea or formula into something computable, reproducible, and adoptable

Workflow: Define → Make Computable → Prove → Validate → Harden

The categories are ordered by cascade severity — an upstream mistake poisons everything below it. Work top-down, and jump straight to a category using this table:

If you are…Start inFirst rule
Starting from a fuzzy propertydef-def-name-the-latent-construct
Worried the ideal is uncomputable / undecidablecomp-comp-do-not-define-metric-as-uncomputable-ideal
Unsure whether you can average or take ratiosmeas-meas-declare-the-scale-type
Claiming the metric behaves a certain wayprop-prop-prove-monotonicity
Getting different numbers between runsdet-det-pin-iteration-and-tie-break-order
Unsure it measures the real thingvalid-valid-discriminant-not-just-loc
Letting an agent optimize the metricgame-game-hard-block-construct-violating-wins
Publishing the metric for othersagg-agg-ship-reference-impl-and-test-vectors

Each reference file is a {category}-{slug}.md containing: WHY it matters, an Incorrect example with the failure annotated, a Correct example with the minimal fix, and a reference. The incorrect/correct examples are metric definitions and procedures, not application code — the contrast is a badly-designed measure versus the fixed one.

Rule Categories by Priority

#CategoryPrefixImpactRules
1Construct Definition & Operationalizationdef-CRITICAL6
2Computability & Tractabilitycomp-CRITICAL7
3Measurement-Theoretic Foundationsmeas-HIGH5
4Proof of Metric Propertiesprop-HIGH6
5Determinism & Reproducibilitydet-HIGH5
6Construct Validity & Calibrationvalid-MEDIUM-HIGH6
7Optimization Safety & Anti-Gaminggame-MEDIUM5
8Aggregation, Reporting & Adoptionagg-LOW-MEDIUM4

See references/_sections.md for the full ordering rationale.

Quick Reference

1. Construct Definition & Operationalization (CRITICAL)

  • def-name-the-latent-construct — Name the unobservable property before writing any formula
  • def-separate-construct-from-proxy — Keep construct, proxy, and their assumed link distinct
  • def-write-falsifiable-operational-definition — Specify the exact procedure that yields the number
  • def-fix-unit-of-analysis — Pin the unit of analysis and the measurement boundary
  • def-anchor-to-the-decision — Attach the decision and act
阅读完整来源 ↗