Skill Builder
v1.0.0技能构建工厂（Meta-Skill Factory）。使用苏格拉底诘问法 + 第一性原理分析，通过多轮对话迭代细化需求，帮助用户从零到一构建一个完整的新技能。当用户说"创造一个技能"、"创建一个 skill"、"帮我设计一个 xxx 技能"、或表达任何需要构建新技能的需求时激活。
⭐ 0· 92·0 current·0 all-time
byBenson@dead1ine
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
The name/description (a Socratic, first-principles skill builder) matches the contents: the files contain question templates, complexity matrices, and an iterative workflow. It does not ask for unrelated binaries, credentials, or system access.
Instruction Scope
SKILL.md contains a clear multi‑round dialog flow for eliciting requirements and producing a SKILL.md and file list. It does not instruct the agent to read local files, environment variables, or hidden endpoints. However the final steps mention installing to the user's workspace and publishing to ClawHub and explicitly say 'tell me your token' — the skill does not declare how tokens should be handled, so there's a risk users may be prompted to provide sensitive tokens directly in chat. Review any requests for credentials carefully.
Install Mechanism
No install spec and no code files — instruction-only. This is the lowest-risk install profile: nothing is downloaded or written by a packaged installer.
Credentials
The skill declares no required environment variables or credentials, which is proportional. Nonetheless the workflow may ask the user for a ClawHub/publishing token or permission to install into a workspace; those are user-supplied secrets and are not required by the skill itself. Treat any token requests as out-of-band actions and only provide limited-scope or temporary tokens if you choose to proceed.
Persistence & Privilege
always is false and the skill does not request persistent presence or attempt to modify other skills. Model invocation is allowed (the platform default) — nothing in the skill claims elevated privileges or system-level changes.
Assessment
This skill is internally consistent and appears benign: it’s an authoring assistant that asks questions and generates SKILL.md drafts. Before using it: (1) do not paste long-lived or high‑privilege tokens into chat — if asked for a publishing/install token, prefer a limited-scope or temporary token and use the platform's secure token input if available; (2) review any generated SKILL.md, file list, and scripts before running them or installing the skill into your workspace; (3) if the skill suggests installing code or running commands, run those actions in a sandboxed environment first; (4) if you plan to publish, prefer performing the publish action via the official UI/CLI with your token entered through a secure prompt rather than sending it in conversation.Like a lobster shell, security has layers — review code before you run it.
latestvk97fpb8kd21n9naccp8a65n9p183g4j9
License
MIT-0
Free to use, modify, and redistribute. No attribution required.