What
Two built-in commands now detect that they are being loaded as part of a skill preload and skip their work: the codebase-analysis prompt skips its expensive preparation, and /insights returns a placeholder instead of generating a report. The report is still produced normally when you run /insights yourself.
Details
- Both branch on the existing skill-preload and read-only-load options rather than a new flag.
- The tool-result formatter gained a matching read-only branch that tells the model nothing was executed.
Evidence
The /insights report is generated only when the command is invoked directly.
Strings lifted out of the shipped bundle, so the claim above can be checked against them.
Names in the bundle/insights
The entry above is what we published on the day. These lines were added later, as Anthropic's own pages caught up, and they sit beside the original rather than replacing it.
Confirmed since
Anthropic's documentation has since written up /insights, on Claude Code changelog.
* Fixed `/insights` failing on Bedrock, Vertex, Foundry, and gateway deployments whose account can't reach the default Opus model by using the session model there insteadchangelog see the edit
One source agreesOne thing we can check says the same as this entry.
Anthropic's documentation agrees
Anthropic's documentation has since written up /insights, on Claude Code changelog.