next
Suggest the next task to work on based on current Phase and worklog state. Use when you want to know what to do next in the project.
Best use case
next is best used when you need a repeatable AI agent workflow instead of a one-off prompt.
Suggest the next task to work on based on current Phase and worklog state. Use when you want to know what to do next in the project.
Teams using next should expect a more consistent output, faster repeated execution, less prompt rewriting.
When to use this skill
- You want a reusable workflow that can be run more than once with consistent structure.
When not to use this skill
- You only need a quick one-off answer and do not need a reusable workflow.
- You cannot install or maintain the underlying files, dependencies, or repository context.
Installation
Claude Code / Cursor / Codex
Manual Installation
- Download SKILL.md from GitHub
- Place it in
.claude/skills/next/SKILL.mdinside your project - Restart your AI agent — it will auto-discover the skill
How next Compares
| Feature / Agent | next | Standard Approach |
|---|---|---|
| Platform Support | Not specified | Limited / Varies |
| Context Awareness | High | Baseline |
| Installation Complexity | Unknown | N/A |
Frequently Asked Questions
What does this skill do?
Suggest the next task to work on based on current Phase and worklog state. Use when you want to know what to do next in the project.
Where can I find the source code?
You can find the source code on GitHub using the link provided at the top of the page.
SKILL.md Source
# Next Task Recommendation Identify the highest-priority next task based on current project state. ## 1. Gather State (Minimal Reads) Read only: - `localdocs/worklog.doing.md` — full (short by design) - `localdocs/worklog.todo.md` — full (short by design) - Project plan if it exists (`localdocs/plan*.md`) — **priority/phase section only**, not the full doc Then run a single directory listing to see what's implemented: ```bash ls -1 src/*/ ``` Do not read `worklog.done.md` — doing + todo + directory state is sufficient. ## 2. Determine Current State From what you read, answer: 1. Is there anything **in progress** (doing file)? → If yes, recommend finishing it first. 2. What does the **plan** say is the current priority? → Use the plan's own phase/priority structure, not hardcoded assumptions. 3. What is **not yet implemented** in `src/`? → Cross-reference plan priorities against actual files. If no plan document exists, infer priorities from todo order and directory gaps. ## 3. Priority Logic 1. **In-progress tasks exist** → recommend completing the current task 2. **Todo has items** → recommend the top item that matches the current phase/priority 3. **Todo is empty** → compare plan priorities against `src/` and suggest what's missing 4. **Phase boundary** → if all current-phase work looks done, ask: "Start next phase?" 5. **Missing tests** → if an implementation file exists with no corresponding test, flag it ## 4. Output Format ``` ## Next Task ### Current State - Phase: [derived from plan or inferred] - In progress: [task or "none"] - Phase completion: [rough estimate, e.g. "3 of ~7 tasks"] ### Recommended Task **[filename or task name]** #### What to do [What needs to be implemented — derived from plan doc, not hardcoded] #### Reference docs [Link to relevant plan section if found] #### Getting started \```bash [command or initial snippet if applicable] \``` #### When done \``` worklog done [task] complete next \``` ``` ## Notes - Derive phase structure and file expectations from the plan document — never assume a fixed structure - If the plan has changed, reflect that; do not rely on stale mental models - Keep reads to: doing + todo + one plan section + one `ls`
Related Skills
worklog
Update worklog files by moving tasks between todo/doing/done states. Use when recording task progress, starting new work, or marking tasks complete. Requires explicit arguments: worklog [done|doing|todo] [description].
template-proposal-review
Review proposal files in localdocs/localdocs/proposals/ and decide whether to apply them to the upstream python-project-template repository. Use when evaluating proposals generated by template-upstream, deciding which ones to accept or reject, and applying accepted ones to the template repo. Triggers on: "proposal 검토", "제안 검토", "template에 반영할지", "proposals 리뷰", "upstream 반영 결정".
template-downstream
Update .claude/, .mcp.json, and .pre-commit-config.yaml from the upstream python-project-template repository into the current project. Only files that exist upstream are updated — project-only files are never deleted. Use when pulling latest tooling, rules, hooks, or skills from the canonical template into this project. Triggers on: "템플릿 최신화", "template sync", "upstream 반영", "template 업데이트".
template-broadcast
Apply template-downstream to all registered downstream projects in bulk, one project at a time. Use when you want to push the latest .claude/, .mcp.json, or .pre-commit-config.yaml to every project at once. Triggers on: "일괄 배포", "전체 프로젝트에 내려보내기", "모든 프로젝트 업데이트", "broadcast template", "bulk sync".
tdd
Test-Driven Development workflow. Use for ALL code changes - features, bug fixes, refactoring. TDD is non-negotiable.
sparks-create
Apply the 13 thinking tools from "Sparks of Genius" (생각의 탄생) to reframe any problem through creative, cross-disciplinary lenses. Use when conventional analysis falls short, when you need a fresh creative angle, or when the problem spans multiple domains.
skill-creator
Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.
rethink-unblock
Break out of a stuck approach by reframing a technical problem through structured thinking frameworks. Use when blocked, going in circles, or when a fresh perspective is needed on architecture, design, or debugging.
refactoring
Refactoring assessment and patterns. Use after tests pass (GREEN phase) to assess improvement opportunities.
python-mcp-expert
Expert assistant for developing Model Context Protocol (MCP) servers in Python. Use when creating, debugging, or optimizing MCP servers with FastMCP, tools, resources, and prompts.
python-conventions
Python coding conventions and guidelines including PEP 8, type hints, docstrings, and testing standards. Automatically applied when working with Python files.
prd
Generate high-quality Product Requirements Documents (PRDs) for software systems and AI-powered features. Includes executive summaries, user stories, technical specifications, and risk analysis.