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.

Turn an agent failure into a test case v1.0.0

Use this resource to turn a real agent failure into a reproducible test case and fix its knowledge

New

  • 405 downloads
  • 244 likes
  • 119 saves

Free

Install with the CLI

In a terminal, from the folder you want:

npx --yes palatekit@latest install agent-eval-regression-case

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

From an observed error, the agent first writes a case that fails, proposes the smallest knowledge change, then checks before and after with the harness commands. Nothing is added without a test that proves the failure.

Mostly picked by

Updated November 2, 2026

Card for the “Turn an agent failure into a test case” skill

Details

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

Similar resources

View all

What you get

  • SKILL.md: arguments, workflow and case rules
  • references/case-template.md and knowledge-change.md: case template and knowledge change

Good fit for

  • An agent gave a wrong answer and you want it not to recur
  • You are extending a project's evaluation harness

Not for

    Tags

    • evaluation
    • agent
    • regression
    • harness
    • knowledge

    About

    You say

    “The agent mixed up two VAT rules: add a case to the harness and correct its knowledge.”

    Card for the “Turn an agent failure into a test case” skill

    Skill published by open-mercato 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 turn a real agent failure into a reproducible test case and fix its knowledge
    3. Review and adjust the result

    Content preview

    Files

    • LICENSE1.1 KB
    • NOTICE-PALATEKIT.md0.5 KB
    • references/agentic-setup.md5.5 KB
    • references/case-template.md3.8 KB
    • references/case-workflow.md3.9 KB
    • references/knowledge-change.md6.2 KB
    • references/owner-selection.md1.5 KB
    • references/report-templates.md1.8 KB
    • references/rules.md6.3 KB
    • SKILL.md5.7 KB

    Origin and licence

    Listed unmodified

    Author
    open-mercato
    Source
    github.com/open-mercato/skills/tree/002bf5cbf9e9ffe4b56120b4f79080cf61df79d4/skills/om-evolve-harness
    Licence
    MIT

    Notice

    « om-evolve-harness » (open-mercato, 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.