@mattshumer_: Your skills are making Opus 5 worse. My prompt fixes that. It goes through each skill and uses a Gauntlet Loop to rewri…
Summary
A free prompt for Claude Code that audits installed skills and uses a Gauntlet Loop to blind-test upgraded versions against originals, upgrading or retiring skills to work better with Opus 5.
View Cached Full Text
Cached at: 08/08/26, 03:05 PM
Your skills are making Opus 5 worse.
My prompt fixes that.
It goes through each skill and uses a Gauntlet Loop to rewrite and blind-test it against the old version, until the new skill + Opus 5 is better by far.
Try it: https://t.co/xzZJwrZG1g
The Opus 5 Skills Upgrade Prompt
Source: https://somethingbig.ai/skills-upgrade ← How to Run a Gauntlet Loop
Free promptClaude Code
Your skills were built for Opus 4.8. Upgrade them for Opus 5.
Skills written for older models over-explain, over-constrain, and patch weaknesses Opus 5 doesn’t have — so they now make itworse. This one prompt audits every skill you’ve got, then puts the ones you pick through a blind-judged gauntlet: upgraded if that honestly wins, retired if it doesn’t. Paste it into Claude Code and hit run.
The prompt
Audit every Claude Code skill currently installed and autonomously upgrade the ones I select through a Skill Upgrade Gauntlet.
First, find and inspect every installed skill across every available scope. Preserve an immutable copy of every original skill and its resources.
Before beginning any upgrades, create a live local HTML dashboard showing the complete skill inventory, what each skill does, where it came from, whether it can be edited, and any dependencies or overlaps you discover. Open the dashboard on my computer and give me its link in Claude Code.
The dashboard is view-only. I…
Unlocked for— copy it and go.
How to run it
01
Open Claude Code
On the machine where your skills live —\.claude/skillsin a project, or~/\.claude/skills.Opus 5 or Fable 5both work.
02
Paste it, pick your skills
It inventories every skill, opens alive dashboard, and asks which ones to put through the gauntlet. Reply with your picks in Claude Code.
03
Let it run
Each pick is blind-benchmarked against the original until it’s honestly better —upgraded or retired, with the evidence on the dashboard. It tells you when it’s done.
**Before you run it:**make sure your skills folder is committed to git or backed up. The prompt tells the agent to keep everything reversible, but your own backup is the one that counts.
Similar Articles
@mattshumer_: For those who want to keep their skills while using Opus 5, here's a trick you can try (let me know how it goes): Write…
Matt Shumer shares a technique for adapting skills to Claude Opus 5: a loop that updates skills, benchmarks them against Opus 4.8 with old skills, and iterates until Opus 5 wins.
@learnwithella: Self-improving Claude Code skills are f*cking ridiculous One loop → 10 test runs, scored against an eval, prompt rewrit…
Claude Code can auto-iterate prompts by running evals, rewriting, and keeping winners, boosting a hook-writer skill from 32/50 to 47/50 overnight.
@addyosmani: Tip: Claude Code has several skill-cleanup commands. Each answers a different question. /skill-doctor → which skills yo…
A tip about skill-cleanup commands in Claude Code, such as /skill-doctor for checking used skills and /doctor for fixing setup issues.
Prompting Claude Opus 5.5
This guide details prompting patterns specific to Claude Opus 5.5, addressing behavioral differences from Claude Opus 5 and providing strategies for effort calibration, API integrations, and unattended agentic tasks.
@mattpocockuk: Prompt of the day: 1. Re-check mattpocock/skills for the v1.3 release. Rename CONTEXT.md to GLOSSARY.md. Check /setup-m…
Matt Pocock shares a 'prompt of the day' for upgrading to the v1.3 release of his mattpocock/skills repository, including renaming CONTEXT.md to GLOSSARY.md and analyzing recent sessions to optimize skill usage.