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.

Run a GEPA learning cycle v1.0.0

Use this resource to optimise a harness policy against a scored corpus

New

  • 526 downloads
  • 182 likes
  • 177 saves

Free

Install with the CLI

In a terminal, from the folder you want:

npx --yes palatekit@latest install gepa-harness-learning-cycle

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 runs metaharness learn on a harness repository and a list of task slices. By default it only estimates the cost; the --run option starts the cycle for real. Only candidates that win on held-out slices are kept.

Mostly picked by

Updated November 2, 2026

Card for the “Run a GEPA learning cycle” skill

Details

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

Similar resources

View all

What you get

  • SKILL.md: prerequisites and algorithm
  • SKILL.md: estimated cost and exit codes

Good fit for

  • A harness fails on one family of tasks
  • You want to price a cycle before running it

Not for

    Tags

    • learning
    • GEPA
    • cost
    • harness

    About

    You say

    “Estimate the cost of a learning cycle on my harness repo, without running it.”

    Card for the “Run a GEPA learning cycle” skill

    Skill published by ruvnet 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 optimise a harness policy against a scored corpus
    3. Review and adjust the result

    Content preview

    Files

    • LICENSE1.1 KB
    • NOTICE-PALATEKIT.md0.5 KB
    • SKILL.md3.2 KB

    Origin and licence

    Listed unmodified

    Author
    ruvnet
    Source
    github.com/ruvnet/ruflo/tree/89b0cd02dd99ea1f63437fde6c05992ec53cc22b/plugins/ruflo-metaharness/skills/harness-learn
    Licence
    MIT

    Notice

    « harness-learn » (ruvnet, 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.