◉CLAUDE LABJP
●CC 2.1.330 — Claude Code 2.1.330 fixes the multi-minute freeze right after compacting a long conversation●SONNET 4.5 — Sonnet 4.5 retires from the Claude API on 11/30, 8 days left. The suggested replacement is claude-sonnet-5-5●MCP 401 — Users report MCP servers that return 401 time out silently instead of showing an error●NEW — New: Desktop creates a worktree for every session? Three places to check●OPUS 4.5 — The 11/24 date for Opus 4.5 is a not-sooner-than floor, not a retirement notice●SCHEMA — Same schema, different shape: compare structured output from the Agent SDK and the Messages API●CC 2.1.330 — Claude Code 2.1.330 fixes the multi-minute freeze right after compacting a long conversation●SONNET 4.5 — Sonnet 4.5 retires from the Claude API on 11/30, 8 days left. The suggested replacement is claude-sonnet-5-5●MCP 401 — Users report MCP servers that return 401 time out silently instead of showing an error●NEW — New: Desktop creates a worktree for every session? Three places to check●OPUS 4.5 — The 11/24 date for Opus 4.5 is a not-sooner-than floor, not a retirement notice●SCHEMA — Same schema, different shape: compare structured output from the Agent SDK and the Messages API
Articles/Claude Code
⟐ Claude Code/2026-08-24Advanced

The same skill lived in two places, and only one copy still had its files

Two scopes held skills with identical names. The SKILL.md files matched down to the sha256, yet one copy was missing the files its own body points to. Here is how I scanned 66 skills and rewrote the detector three times to get from four hits to two real ones.

Claude Code291skill designnamespacesintegrity checksoperations33

✦ Premium Article

One line in the August 23 release notes caught my eye. A bundled skill alias was returning "Unknown command" when a user-scoped or project-scoped skill of the same name shadowed it, and that had been fixed.

I assumed it did not apply to me. I had been careful about naming.

I counted anyway. Forty-two skills on the project side, twenty-four on the user side, sixty-six in total. Five of them existed under the same name in both places.

That much was still within expectations. What followed was not. All five pairs had SKILL.md files that matched down to the sha256 — and two of those pairs held a different number of bundled files.

A name collision is not, by itself, an incident

Which copy wins resolution is a separate question, and I wrote about it in Nothing came back from the skill I called — pinning down name resolution before launch. The conclusion there was to build a table of effective names before starting.

What I ran into this time sits one layer beneath that. Even when the effective name is settled, the entity that wins is not guaranteed to be complete.

When the same skill sits in two places, the natural first move is to compare the bodies. That is what I did.

diff .agents/skills/seo-audit/SKILL.md ~/.claude/skills/seo-audit/SKILL.md
# no output (zero differences)

Zero differences. Both 10,688 bytes. I closed the terminal feeling reassured.

Then I realized I had never looked at the directories themselves.

What was missing was not the body, but the files the body names

Here are the same two directories, listed file by file.

(cd .agents/skills/seo-audit && find . -type f | sort)
# ./SKILL.md
# ./evals/evals.json
# ./references/ai-writing-detection.md
 
(cd ~/.claude/skills/seo-audit && find . -type f | sort)
# ./SKILL.md

Identical bodies. Three files against one.

And the body of SKILL.md contains this line:

references/ai-writing-detection.md

So if the user-scoped copy wins, the skill goes looking for a file that is not there. Because the two SKILL.md files match exactly, neither diff nor sha256sum will ever surface this.

The missing references/ai-writing-detection.md is 6,528 bytes — roughly sixty percent of the size of SKILL.md itself, referenced but absent.

What you compareProject copyUser copyDoes it show?
SKILL.md contents10,688 bytes10,688 bytes (identical)No
SKILL.md sha256matchmatchNo
Files in the directory31Yes
Referenced files presentall presentone absentYes

I had been operating on the top two rows alone.

✦

Thank you for reading this far.

Continue Reading

What follows includes implementation code, benchmarks, and practical content we hope you'll find useful. This site runs without ads — server and development costs are supported entirely by members like you. If it's been helpful, we'd be truly grateful for your support.

WHAT YOU'LL LEARN
✦You will be able to check, in your own setup, whether the skill that actually wins name resolution is the complete one
✦You will be able to catch a failure mode that survives a byte-for-byte diff, using a single check on whether referenced files exist
✦You will be able to separate false positives from false negatives instead of trusting the count your detector prints
Secure payment via Stripe · Cancel anytime
✦

Unlock This Article

Get full access to the rest of this article. Buy once, read anytime. This site is ad-free — your support goes directly toward keeping it running.

or
Unlock all articles with Membership →
Share

Thank You for Reading

Claude Lab is ad-free, supported entirely by members like you. We publish practical guides daily with implementation code, benchmarks, and production-ready patterns. If you've found it useful, we'd love to have you on board.

  • ✦Copy-paste ready implementation code
  • ✦New advanced guides published daily
  • ✦$5/mo or $10 for lifetime access
View Membership →

Related Articles

⟐ Claude Code2026-11-06
Claude Code Shipped 21 Releases in Four Weeks. I Counted Before Deciding How Often to Update
Should you check Claude Code updates every day or once a week? I counted npm's publish history with a standard-library Python script and used the last 28 days of counts and gaps to set a rhythm.
⟐ Claude Code2026-10-24
The Settings File I Kept Saying Yes To Had a Rule That Let Any Python Through — Auditing Allow Rules After the 2.1.309 Changelog
Claude Code 2.1.309 is full of permission-check and sandbox fixes. Here is a script that audits the allow rules in settings.local.json, and how to fix a prefix wildcard on an interpreter.
⟐ Claude Code2026-09-22
Taking Inventory of Worktrees That Were Supposed to Be Gone
Isolated worktrees my subagents created were never folded away. Here is what I measured in git 2.34.1 about when remove succeeds, the branches that outlive the trees, the 12 MB per-tree difference, and how I decide who owns cleanup.