[DOCS] Missing pattern for creating "General Purpose" subagents with Skills access

Documentation Type

Missing documentation (feature not documented)

Documentation Location

https://code.claude.com/docs/en/sub-agents

Section/Topic

The "Skills and subagents" section, specifically the note box regarding built-in agents.

Current Documentation

"Built-in agents (Explore, Plan, general-purpose) do not have access to your Skills. Only custom subagents you define in .claude/agents/ with an explicit skills field can use Skills."

What's Wrong or Missing?

The documentation identifies a significant limitation—that the capable "general-purpose" built-in agent cannot use Skills—but fails to provide the solution for this common workflow.

Users logically want the broad reasoning capabilities of the general-purpose agent combined with their specific project Skills. The current text implies this combination is impossible or requires the user to guess how to reconstruct a "general-purpose" agent from scratch using a custom definition.

Suggested Improvement

Please add a "Best Practice" or "Workaround" example immediately following that note, showing how to define a custom agent that effectively replicates the general-purpose agent but with skills enabled.

Suggested text addition:

Tip: Creating a General Purpose Agent with Skills If you need the capabilities of the general-purpose agent but with access to your skills, create a custom agent that inherits all tools and models, and explicitly lists your skills: ``yaml # .claude/agents/skilled-generalist.md --- name: skilled-generalist description: A general purpose agent that can use all tools and project skills. Use for complex multi-step tasks. model: inherit # Omit 'tools' to inherit all tools (like the built-in general-purpose agent) skills: code-review, database-helper, api-docs --- You are a capable general-purpose assistant. Use your available skills and tools to complete complex tasks. ``

Impact

High - Prevents users from using a feature

Additional Context

This logic gap creates a workflow failure where users define high-value Skills but find them inaccessible during complex, multi-step operations (which are usually handled by the general-purpose agent). Providing the configuration pattern above solves this blocking issue.