ClawHub

Skill Builder / Creator

v1.0.5

Create high-quality skills with modular structure, progressive disclosure, and token-efficient design.

⭐ 15· 5.5k·79 current·85 all-time
byIván@ivangdavila
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Benign
high confidence
✓
Purpose & Capability
The name/description promise (help authors create skills, progressive disclosure, token-efficient design) aligns with the actual artifacts: SKILL.md, setup.md, patterns.md, memory-template.md. It requests no environment variables, binaries, or install steps — all coherent for a documentation/authoring assistant.
✓
Instruction Scope
Runtime instructions are limited to authoring guidance, structure templates, and a policy to 'ask user first' before creating files. The only filesystem interaction suggested is creating a project folder under the user's home directory (explicitly gated by 'ask user first'). No external endpoints, credentials, or opaque data-collection steps are instructed.
✓
Install Mechanism
There is no install spec and no code files — this is instruction-only, which minimizes risk. The README references a 'clawhub install' command for related skills, but that is advisory and not an automated install step in this skill.
✓
Credentials
The skill declares no required environment variables, credentials, or config paths. The SKILL.md and auxiliaries do not instruct the agent to read secrets or unrelated environment variables.
✓
Persistence & Privilege
always is false and disable-model-invocation is the platform default (false). The skill does not request persistent system-wide changes or access to other skills' configs. It documents a user-confirmation policy for any file creation.
Assessment
This skill is instruction-only and internally consistent with its stated purpose. Before using: (1) confirm the agent asks you before creating any files in your home directory; (2) do not approve running external install commands (e.g., 'clawhub install ...') or shell commands that download code unless you trust the source; (3) verify the skill source/homepage if you plan to follow related-tool install instructions; and (4) if you want stronger guarantees, disable autonomous invocation for the agent or require explicit confirmation before any filesystem or network action.

Like a lobster shell, security has layers — review code before you run it.

latestvk977281xt78evzwk6c0gfnrx5h81v7n6

License

MIT-0
Free to use, modify, and redistribute. No attribution required.
Termshttps://spdx.org/licenses/MIT-0.html

Runtime requirements

🛠️ Clawdis
OSLinux · macOS · Windows

Comments