[Issue #75] Create /create-capability command for scaffolding capability sets #83

Merged
HugoNijhuis merged 1 commits from issue-75-create-capability-command into main 2026-01-10 18:22:50 +00:00
Owner

Summary

Adds a new /create-capability command that helps users create new capabilities for the architecture repository. A capability may be a single component (skill, command, or agent) or a cohesive set working together.

Changes

  • Created commands/create-capability.md with guided workflow for:
    • Analyzing user descriptions to understand capability needs
    • Recommending appropriate component combinations (skill only, command only, command+skill, command+agent, or full set)
    • Gathering information for each component type
    • Selecting appropriate models (haiku, sonnet, opus) based on task complexity
    • Generating files from templates with proper inter-references
    • Presenting preview for approval before file creation
    • Creating files in correct locations

Acceptance Criteria

  • Command created at commands/create-capability.md
  • References capability-writing skill via @~/.claude/skills/capability-writing/SKILL.md
  • Determines appropriate component set based on description
  • Generates cohesive files with proper inter-references
  • Selects appropriate models for each component
  • Shows preview before creating files
  • Creates files in correct locations

Closes #75

## Summary Adds a new `/create-capability` command that helps users create new capabilities for the architecture repository. A capability may be a single component (skill, command, or agent) or a cohesive set working together. ## Changes - Created `commands/create-capability.md` with guided workflow for: - Analyzing user descriptions to understand capability needs - Recommending appropriate component combinations (skill only, command only, command+skill, command+agent, or full set) - Gathering information for each component type - Selecting appropriate models (haiku, sonnet, opus) based on task complexity - Generating files from templates with proper inter-references - Presenting preview for approval before file creation - Creating files in correct locations ## Acceptance Criteria - [x] Command created at `commands/create-capability.md` - [x] References capability-writing skill via `@~/.claude/skills/capability-writing/SKILL.md` - [x] Determines appropriate component set based on description - [x] Generates cohesive files with proper inter-references - [x] Selects appropriate models for each component - [x] Shows preview before creating files - [x] Creates files in correct locations Closes #75
HugoNijhuis added 1 commit 2026-01-10 18:19:59 +00:00
Introduces a new command that guides users through creating capabilities
for the architecture repository. The command analyzes user descriptions,
recommends appropriate component combinations (skill, command, agent),
gathers necessary information, generates files from templates, and presents
them for approval before creation.

Closes #75

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Author
Owner

AI Code Review

This is an automated review generated by the code-reviewer agent.

Summary

The /create-capability command provides a structured workflow for scaffolding new capabilities in the architecture repository. The implementation follows established command patterns and integrates well with the capability-writing skill.

Findings

Code Quality

  • Well-structured workflow with clear phases (understand, determine, gather, generate, preview, create)
  • The component selection matrix (lines 19-26) provides clear guidance on when to use each pattern
  • Model selection guidance (lines 67-70) is practical and actionable
  • Good use of tables for presenting information clearly
  • The inter-reference requirements (lines 76-78) are well documented

Potential Bugs

  • Line 8: The command references "" in line 11 but the argument-hint shows "" - ensure this aligns with how arguments are passed to commands in the framework
  • Lines 76-78: The inter-reference instructions specify different formats (@ path for commands vs. names-only for agents) - this is correct based on the architecture but could benefit from validation
  • No error handling specified for cases where the capability-writing skill is not available

Security Concerns

  • No security concerns identified
  • Command creates files but requires approval first (line 81), which is appropriate
  • No sensitive data handling

Style Notes

  • Consistent with other command files in the repository
  • Frontmatter follows the established pattern with description and argument-hint
  • Clear section headers and formatting
  • The workflow steps are well-numbered and easy to follow

Test Coverage

  • This is a markdown command file, so traditional unit tests don't apply
  • Consider adding:
    • Example walkthrough in documentation showing the full flow
    • Reference to issue #75 for acceptance criteria validation
    • Examples of each component pattern (skill-only, command-only, full set)

Verdict

LGTM

The command is well-designed with clear guidance for users and proper integration with the capability-writing skill. The structured approach ensures consistent capability creation. Minor suggestions above are for enhancement but not blocking.

## AI Code Review > This is an automated review generated by the code-reviewer agent. ### Summary The /create-capability command provides a structured workflow for scaffolding new capabilities in the architecture repository. The implementation follows established command patterns and integrates well with the capability-writing skill. ### Findings #### Code Quality - Well-structured workflow with clear phases (understand, determine, gather, generate, preview, create) - The component selection matrix (lines 19-26) provides clear guidance on when to use each pattern - Model selection guidance (lines 67-70) is practical and actionable - Good use of tables for presenting information clearly - The inter-reference requirements (lines 76-78) are well documented #### Potential Bugs - Line 8: The command references "" in line 11 but the argument-hint shows "<description>" - ensure this aligns with how arguments are passed to commands in the framework - Lines 76-78: The inter-reference instructions specify different formats (@ path for commands vs. names-only for agents) - this is correct based on the architecture but could benefit from validation - No error handling specified for cases where the capability-writing skill is not available #### Security Concerns - No security concerns identified - Command creates files but requires approval first (line 81), which is appropriate - No sensitive data handling #### Style Notes - Consistent with other command files in the repository - Frontmatter follows the established pattern with description and argument-hint - Clear section headers and formatting - The workflow steps are well-numbered and easy to follow #### Test Coverage - This is a markdown command file, so traditional unit tests don't apply - Consider adding: - Example walkthrough in documentation showing the full flow - Reference to issue #75 for acceptance criteria validation - Examples of each component pattern (skill-only, command-only, full set) ### Verdict **LGTM** The command is well-designed with clear guidance for users and proper integration with the capability-writing skill. The structured approach ensures consistent capability creation. Minor suggestions above are for enhancement but not blocking.
HugoNijhuis merged commit d5deccde82 into main 2026-01-10 18:22:50 +00:00
Sign in to join this conversation.
No Reviewers
No Label
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: flowmade-one/architecture#83