Task Planner
v0.1.0任务规划与执行引导工具。当用户提出需要多步骤完成的复杂任务、请求帮忙规划工作、说"帮我规划"、"拆解任务"、"怎么完成这个"时触发。也适用于用户描述了一个大目标但不确定从哪里开始的场景。即使用户没有明确说"规划"，只要任务明显涉及3个以上步骤、多个依赖关系、或需要分阶段完成，都应该使用此 skill。覆盖场景包括...
⭐ 0· 74·0 current·0 all-time
by@jiangrb
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
The skill is an instruction-only 'task planner' and only references planning and execution tools (TaskCreate, TaskUpdate, TaskList) and common authoring/editing tools for programming tasks. It does not request unrelated binaries, environment variables, or installs—this matches the stated purpose.
Instruction Scope
The SKILL.md stays focused on planning and guided execution. It instructs using TaskCreate/TaskUpdate/TaskList to persist and manage progress and, for programming tasks, to use editor/Bash tools to perform work. This is reasonable for a planner, but these tool calls imply the ability to create/modify tasks and potentially read/write files or run commands—so the operator should confirm the platform's semantics and consent model before allowing autonomous execution.
Install Mechanism
No install specification or code files are present (instruction-only). No downloads, package installs, or binaries are required—lowest-risk install profile.
Credentials
The skill declares no required environment variables, credentials, or config paths. The SKILL.md does not instruct access to secrets or unrelated environment values.
Persistence & Privilege
always is false and the skill is user-invocable. It requests normal task-list persistence via TaskCreate/TaskUpdate, which is proportionate to a planner. Note: the platform default allows the model to invoke skills autonomously; that is not flagged here but users should be aware of it.
Assessment
This skill appears coherent and appropriate for planning and guiding multi-step tasks. Before enabling it to execute actions, confirm: (1) what the platform's TaskCreate/TaskUpdate/TaskList tools store and where (privacy/persistence), (2) whether the agent will prompt you before reading/writing files or running bash commands, and (3) whether you want to permit autonomous execution. If you prefer safer operation, restrict the agent from running file-modifying commands or require explicit human confirmation before any destructive or privileged action.Like a lobster shell, security has layers — review code before you run it.
latestvk977ba061mpmzpzswetc44yg8d84czen
License
MIT-0
Free to use, modify, and redistribute. No attribution required.