Three new skills replacing the deprecated commit-commands plugin:
- git-commit: Convention-aware commit creation with logical commit
splitting at the file level when changes have distinct concerns
- git-commit-push-pr: Adaptive PR descriptions that scale in depth with
change complexity, conditional test plans, and GitHub-safe formatting
- git-clean-gone-branches: Fetch-first branch cleanup with user
confirmation and worktree handling, using a shell script for reliable
parsing
All three use `command git` to bypass RTK proxies and shell aliases.
The skill generated outdated output — referenced deprecated tools
(TodoWrite, MultiEdit), project-specific integrations, and used
command-style templates that no longer match current skill conventions.
The plugin ecosystem has better skill-authoring tooling elsewhere.
- Replace deprecated Commands section with proper Skills sections
- Move Skills before Agents — skills are the entry points, agents are implementation details
- Add Core Workflow section with all ce:* skills front and center
- Add Git Workflow section for new git-commit/push/pr skills
- Add Workflow Utilities section absorbing former Utility Commands
- Add Beta/Experimental section for lfg and slfg
- Add note that agents are invoked by skills, not directly by users
- Update skill count from 40+ to 43
Beta skills now use disable-model-invocation: true to prevent accidental
auto-triggering. Descriptions written as future stable descriptions with
[BETA] prefix for clean promotion. Updated solutions doc and AGENTS.md
promotion checklist to include removing the field.
- Beta plans use -beta-plan.md suffix to avoid clobbering stable plans
- Fix internal references in beta skills to use beta names consistently
- Add beta skills section to AGENTS.md with promotion checklist
Create separate beta skills instead of gating existing ones. Stable
ce:plan and deepen-plan are restored to main versions. Beta skills
reference each other and work standalone outside lfg/slfg orchestration.
- Surface deferred implementation questions and scope boundaries
- Use per-unit Patterns and Verification fields for task execution
- Add execution strategy: inline, serial subagents, or parallel
- Reframe Swarm Mode as Agent Teams with opt-in requirement
- Make tool references platform-agnostic
- Remove plan checkbox editing during execution