What problem does it solve?
This Skill tackles the complexity of managing multi-branch Git workflows, dynamically loading necessary tools, and maintaining context across different development phases. It prevents manual errors, ensures consistency, and reduces cognitive load for developers, letting them focus on coding.
Core Features & Use Cases
- Dynamic Skill Loading: Automatically identifies the current workflow phase (e.g., planning, feature development, release) and loads only the relevant skills, keeping your environment lean and efficient.
- Context-Aware Guidance: Uses
TODO_*.md files to track progress and provides intelligent "next step" prompts, ensuring you always know what to do next.
- Automated State Management: Monitors AI context usage, automatically saves workflow state to
TODO_*.md at critical checkpoints, and guides you to optimize context, preventing token overflow and data loss.
- Use Case: As a developer, simply ask "next step?" and the orchestrator will guide you through creating planning documents, setting up a feature worktree, generating specifications, implementing code, running quality checks, and managing releases, all while handling the underlying tool invocation and state updates.
Quick Start
Example: Ask for the next step in your workflow
(This is a conceptual interaction, the orchestrator would
interpret a natural language prompt like "next step?")
The orchestrator would then call the appropriate script, e.g.:
python .claude/skills/bmad-planner/scripts/create_planning.py \
my-feature my-username
