Every premium and members-only resource for $9.99/month, as a no-commitment subscription. See the plans

PalateKit

Sign in
Menu

Request a resource

Describe the resource you are missing in one sentence. We will email you when it is published.

0 / 280

A page, a video or an image online.

Unlock PalateKit

Unlock everything

Plan

Paying means accepting our terms of sale.

Secure payment by Stripe. Receipts and invoices sent by email.

Welcome back

Sign in to install, download or save this resource. A free account is enough for free resources.

Calibrating a classifier's probabilities v1.0.0

Use this resource to check how far a model's probabilities can be trusted and set its thresholds

New

  • 479 downloads
  • 122 likes
  • 111 saves

Free

Install with the CLI

In a terminal, from the folder you want:

npx --yes palatekit@latest install classifier-probability-calibration

With no option, the skill goes to ~/.agents/skills/; add --agent claude-code (or codex, cursor…) to target an agent.

First time: connect the CLI with

npx --yes palatekit@latest login

The agent builds a labelled corpus, measures calibration (ECE, Brier, risk and coverage), then fits temperature or Platt scaling. You get thresholds backed by measurement and a correction for option-order bias.

Mostly picked by

Updated November 2, 2026

Card for the “Calibrating a classifier's probabilities” skill

Details

Works in
Install guides
Generates with
Your agent's model
Licence
MIT

Similar resources

View all

What you get

  • SKILL.md: four-step method and calibration measures
  • SKILL.md: temperature or Platt fitting, order bias and per-action thresholds

Good fit for

  • You want to know whether a classifier's confidence can be trusted
  • A model must abstain below a given probability threshold

Not for

    Tags

    • calibration
    • probabilities
    • threshold
    • classifier

    About

    You say

    “Tell me whether my classifier's confidence scores are reliable and from which threshold it should abstain.”

    Card for the “Calibrating a classifier's probabilities” skill

    Skill published by claude-dev-suite under the MIT license, delivered in its original version.

    Show more

    How to use it

    1. Install the skill in your agent's skills folder
    2. Ask your agent to check how far a model's probabilities can be trusted and set its thresholds
    3. Review and adjust the result

    Content preview

    Files

    • LICENSE1.1 KB
    • NOTICE-PALATEKIT.md0.5 KB
    • SKILL.md16.4 KB

    Origin and licence

    Listed unmodified

    Author
    claude-dev-suite
    Source
    github.com/claude-dev-suite/claude-dev-suite/tree/e10ab644ba43abe2abc923399226df4af7337214/skills/ai-integration/decision-model-calibration
    Licence
    MIT

    Notice

    « decision-model-calibration » (claude-dev-suite, MIT) livré dans sa version d'origine : fiche en français et en anglais par PalateKit.

    User reviews

    –

    No ratings yet

    Loading reviews…

    Delete your review?

    Your rating and comment will be removed from the page.