ClawHub

Ollivanders Agent Shop

v0.1.4

Design and land persona-driven agents by matching role, work identity, cognitive structure, personality grounding, identity-binding, and file skeleton into o...

⭐ 0· 74·0 current·0 all-time
by@child2d
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 (persona-driven agent design and landing) matches the actual content: a methodology, templates, and step-by-step instructions for designing and scaffolding agents. There are no unrelated required binaries, environment variables, or config paths declared.
ℹ
Instruction Scope
The SKILL.md explicitly guides the agent to 'inspect existing agent structure if relevant' and to create files only after user confirmation. This is coherent with its purpose (reviewing and landing an agent), but it implies the agent may read project files or directory structure when the user indicates that's relevant. The skill also instructs generating config snippets for OpenClaw (openclaw.json) rather than silently editing runtime config, which is the safer default.
✓
Install Mechanism
No install spec is provided and there are no code files to execute — this is instruction-only, which is the lowest-risk install model.
✓
Credentials
No environment variables, credentials, or config paths are requested. The skill does not declare or require secrets and therefore does not ask for access beyond ordinary file read/write operations that would be needed to scaffold an agent.
✓
Persistence & Privilege
always is false and the skill does not request persistent presence or to modify other skills' configurations. It may produce file scaffolding and config snippets, but only after design confirmation; it explicitly warns against silently mutating runtime config.
Assessment
This skill is coherent and instruction-only, but be aware it will ask to inspect your project/agent files when that is relevant and can generate filesystem scaffolding and OpenClaw config snippets. Before allowing any file reads or writes: (1) confirm which directories/files the skill should access, (2) review and apply any generated openclaw.json snippets yourself (do not allow silent edits), and (3) test the generated agent in a sandbox or non-production workspace if it will interact with sensitive data or systems.

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

latestvk9763s66emqc1zvw6qyfzkt5x183v2h7

License

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

Comments