You'll noticeTier: how much it should matter to you
1Useful: my rating, 1 to 5
1Signal: worth watching, 1 to 5
PermissionsArea: what it touches
ImprovementsKind: in v2.1.284,
ImprovementsSection of the release
What
A skill is a set of instructions a plugin adds to Claude Code, and its allowed-tools list grants it tools without asking you. When an administrator sets allowManagedPermissionRulesOnly, Claude Code decides whether to honour a plugin skill's allowed-tools. Before, that decision looked only at the plugin's id. It now also looks at:
Whether the plugin is officially attested
Which marketplace it came from
How it was resolved from npm, if it was
Its home location
Plugins can be trusted by attestation or by marketplace source. The same details are also kept with the plugin's commands.
Why
In tightly managed organisations, officially attested plugins or plugins from trusted marketplaces can keep the tool grants their skills rely on.
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 sinceAnthropic's documentation has since written up allowManagedPermissionRulesOnly, on Extend Claude with skills.* The frontmatter applies in every kind of session, so an `allowed-tools` grant goes through the normal [permission flow](/docs/en/permissions). If your organization sets `allowManagedPermissionRulesOnly`, the grant [doesn't apply](#when-o…skillssee the edit
Added sinceA small documentation edit on Extend agents with skills touched a line naming allowManagedPermissionRulesOnly after this was published.In SDK sessions, you can pre-approve tools for a project or personal skill with the skill's [`allowed-tools`](/docs/en/skills#pre-approve-tools-for-a-skill) frontmatter or with the `allowedTools` option (`allowed_tools` in Python) in your …agent-sdk/skillssee the edit
How sure we are
Two sources agreeTwo things we can check say the same as this entry.
Anthropic's documentation agreesA small documentation edit on Extend agents with skills touched a line naming allowManagedPermissionRulesOnly after this was published.
Anthropic's release notes agreeFixed plugins from marketplaces, claude.ai and npm pre-approving their own tools via allowed-tools under managed…
