Specifying And Planning
Productivityspecifying-and-planningSpecifying And Planning is an Agent Skill for productivity published in the majiayu000/claude-skill-registry repository on GitHub: a SKILL.md instruction file an AI agent can load to do this task.
Multi-mode skill for project planning workflows -- extract requirements from specs, break into tasks, create GitHub issues, set up project boards, or generate agent team configurations. Use when planning a project, extracting requirements, breaking down tasks, creating issues from tasks, setting up GitHub projects, or generating team plans.
How to use this skill
Bring this guide into your coding agent with a prompt tailored to the tool you use.
- Open your project in Codex.
- Copy and paste the prompt below.
- Review files and risks before approving any action.
Inspect the Agent Skill at https://github.com/majiayu000/claude-skill-registry/blob/HEAD/skills/planning/specifying-and-planning/SKILL.md and all supporting files. Report provenance, permissions, scripts, dependencies, and security risks. Propose an exact installation into .agents/skills/specifying-and-planning, but wait for my approval before writing files or running scripts.
Copying this prompt does not install or run the skill. Review third-party files before use.
Specifying and Planning
Multi-mode skill for the full project planning pipeline:
spec -> requirements -> tasks -> issues -> team
Modes
spec-to-requirements
Trigger: "extract requirements", "analyze spec", references to spec.md
Analyze a spec file and produce structured requirements.md:
- Functional Requirements: Features, user interactions, data, integrations
- Non-Functional Requirements: Performance, security, usability, reliability
- Requirement Dependencies: Prerequisites, interdependencies, optional links
Each requirement gets clear acceptance criteria.
requirements-to-tasks
Trigger: "break into tasks", "create task breakdown", references to requirements.md
Convert requirements into implementable tasks.md:
- Task description -- specific functionality to implement
- Acceptance criteria -- how to verify completion
- Implementation approach -- TDD, refactoring, new feature
- Required components -- files, modules, systems needing changes
- Test requirements -- what tests to write/update
- Dependencies -- prerequisite tasks, interdependent tasks
Focus on functional decomposition. No timeline or phases -- clean technical breakdown.
tasks-to-issues
Trigger: "create issues", "convert tasks to issues", references to tasks.md
Create GitHub issues from task breakdown:
gh issue create --title "Clear title" --body "$(cat <<'EOF'
## Description
Task description
## Acceptance Criteria
- [ ] Criterion 1
- [ ] Criterion 2
## Implementation Approach
Technical strategy
## Dependencies
- Requires #<issue>
EOF
)"
Apply labels: component (frontend, backend), type (feature, refactor), complexity (small, medium, large).
setup-github-project
Trigger: "set up project", "create project board", references to issues.md
Create complete GitHub project infrastructure:
- Parse issues file for labels, milestones, relationships
- Create labels with smart color coding
- Set up milestones with due dates
- Create project board with custom fields
- Generate all issues with proper labels and milestones
- Link issue dependencies
tasks-to-team
Trigger: "generate team plan", "create agent team", "parallelize work"
Design agent team configuration from tasks:
- Read tasks for dependencies and scope
- Identify parallel work groups (no cross-dependencies)
- Design teammate roles with distinct file ownership
- Define coordination strategy and phase gates
- Write
docs/team.mdwith kickoff prompt
Target 3-5 teammates, 5-6 tasks per teammate. Prefer fewer focused teammates over many scattered ones.
General Principles
- No artificial timelines or phases (except team coordination)
- Focus on deliverables and dependencies
- Each requirement/task/issue should be independently verifiable
- Use TDD approach in task descriptions
- Cache project information in CLAUDE.md