vocion/vocion-core/measure-before-you-make

Claude Code and agent skill by vocion/vocion-core

Trust tier 3
  • community
Source check pending

Checking that the SKILL.md file is still in the repository.

The standing rule that the number a piece is judged on, its baseline, its source and its attribution model are fixed before the work starts — and that "we cannot measure this" is a complete and acceptable answer, written down rather than worked around with a proxy nobody agreed to. Attached to briefing, ranking and reading so the same rule holds at all three.

Install

npx vskill@latest install vocion/vocion-core/measure-before-you-make

SKILL.md

The same SKILL.md was published earlier as vocion/vocion-core/measure-before-you-make.

Measure before you make

Chosen afterwards, any piece looks like a success under some number. Chosen first, it can fail honestly — and honest failure is the only thing the loop learns from.

Four things, fixed in advance

  1. The number, in the workspace's own vocabulary.
  2. The baseline — where it stood when the brief was written. Taken before,

or the brief is not ready. A baseline reconstructed afterwards is a baseline chosen to suit the result.

  1. The source — the system the reading comes from, named so anybody can take

it again. A measure with no source is not a measure.

  1. The attribution model — first touch, last touch, multi-touch, holdout,

none. Identical data produces different winners under different models, and a reading that does not say which cannot be argued with.

And the date the reading is due, so a piece measured after three days is never silently compared with one measured after ninety.

"We cannot measure this" is a real answer

Where no source exists, or the volume is too small for any reading to mean anything, say so at brief time and expect the verdict unmeasurable. Name the instrument that is missing and what it would cost to have it.

What not to do: substitute a proxy nobody agreed to because it happens to be available. A number that can be read but does not answer the question is worse than a gap, because it will be reported, argued over and acted on.

Never a rate on four events

Volume decides whether a ratio means anything. Below the threshold where it does, report the counts. A conversion rate on four events is a story about four people wearing the clothes of a statistic.

Establish normal variation before calling a change

Know what the number does when nothing happens. A movement inside that range is no_effect, and reporting it as a win is how a team spends a year repeating something that never worked.

Unmeasured is reported, never omitted

A channel or a piece with no reading is listed as unread, with the reason. An omitted one reads as one with no return, which is a claim nobody made and nobody checked.

Security

  • Security scanPassed, 90 of 100Pattern checks on the skill's files.
  • Model reviewPassed

Quality evaluation

  • Task evaluationNot runNo evaluation yet. Run one in Skill Studio.

Source

Popularity

Repository stars
2
Repository forks
1
7-day trend score
10

Stars and forks are for the whole GitHub repository, not this skill alone. Not refreshed yet.

Works with

Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.

  • Claude Code
  • Cursor
  • GitHub Copilot
  • Windsurf
  • Codex
  • Gemini CLI
  • Cline
  • Amp
  • Roo Code
  • Goose
  • Continue
  • Kilo Code
  • Junie
  • Trae
  • OpenHands
  • Qwen Code
  • AI Agent
  • Tabnine Chat
  • Sourcegraph Cody
  • Amazon Q
  • Firebase Genkit
  • Sweep
  • Devin
  • Phind
  • Mentat
  • Claude Desktop
  • Aider
  • GPT Pilot
  • Plandex
  • bolt.new
  • v0
  • Replit Agent
  • Anse
  • TypingMind
  • BetterChatGPT
  • LibreChat
  • Warp AI
  • Pieces
  • Fig AI
Embed badge
vskill VERIFIED badge
[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/vocion/vocion-core/measure-before-you-make/badge)](https://verified-skill.com/skills/vocion/vocion-core/measure-before-you-make)

Something wrong with this skill or its page? Tell us.