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

PalateKit

Menu

Request a resource

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

0 / 280

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.

Diagnose an agent failure instead of retrying v1.0.0

Use this resource to understand why an agent loops, burns tokens without progress or drifts from its task

New

  • 349 downloads
  • 204 likes
  • 116 saves

Free

Install with the CLI

In a terminal, from the folder you want:

npx --yes palatekit@latest install agent-failure-self-debugging

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 a four-phase loop: failure capture, root-cause diagnosis, contained recovery and report. You get a reproducible diagnosis and a clear decision on what to do next, before any human escalation.

Mostly picked by

Updated November 2, 2026

Card for the “Diagnose an agent failure instead of retrying” skill

Details

Generates with
Your agent's model
Licence
MIT

Similar resources

View all

What you get

  • SKILL.md: failure capture, root-cause diagnosis and scope boundaries
  • SKILL.md: contained recovery, introspection report and recovery heuristics

Good fit for

  • An agent hits its tool-call limit without producing anything
  • Output quality drops during a long session

Not for

    Tags

    • agent
    • debugging
    • diagnosis
    • loop
    • report

    About

    You say

    “This agent has looped on the same tool for an hour: diagnose the cause and propose a recovery.”

    Card for the “Diagnose an agent failure instead of retrying” skill

    Skill published by affaan-m 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 understand why an agent loops, burns tokens without progress or drifts from its task
    3. Review and adjust the result

    Content preview

    Files

    • LICENSE1 KB
    • NOTICE-PALATEKIT.md0.5 KB
    • SKILL.md5.5 KB

    Origin and licence

    Listed unmodified

    Author
    affaan-m
    Source
    github.com/affaan-m/ecc/tree/25278e7d7b6cf6e6a4e7a6c8e10f14ff22f0eb5b/skills/agent-introspection-debugging
    Licence
    MIT

    Notice

    « agent-introspection-debugging » (affaan-m, 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.