ClawHub

Skill Finder (Find ClawHub skills + Search Skills.sh)

v1.1.5

Find, compare, and install agent skills across ClawHub and Skills.sh when the user needs new capabilities, better workflows, stronger tools, or safer alterna...

⭐ 29· 10.4k·100 current·106 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 skill declares only one runtime dependency (npx) and uses it to run CLIs (clawhub, skills) that are directly relevant to searching and installing skills. No unrelated binaries, environment variables, or config paths are requested.
ℹ
Instruction Scope
The SKILL.md stays on-topic (search, evaluate, recommend, and optionally install skills) and keeps local state under ~/skill-finder. Two items deserve attention: (1) the skill instructs agents to read/write memory.md under ~/skill-finder (expected) and (2) it recommends adding a line to AGENTS.md to steer routing — this modifies workspace routing behavior and could increase how often the skill is loaded. Both are coherent for a 'finder' skill but the AGENTS.md change is broader than simple local state and should be applied only with user consent/backups.
ℹ
Install Mechanism
This is an instruction-only skill (no install spec). It relies on npx to invoke third-party CLIs (clawhub, skills). That is proportional to the purpose, but npx fetches and executes code from npm at runtime, which is inherently higher-risk than pure local-only commands; users should review candidate packages before running install commands.
✓
Credentials
No environment variables, secrets, or external config paths are requested. Local storage is confined to ~/skill-finder and the skill explicitly documents what it stores and what it will not write.
✓
Persistence & Privilege
The skill is not always-on and does not request elevated privileges. It suggests adding a routing line to AGENTS.md to make loading easier; this is a configurable change and not required for basic operation. Autonomous invocation is allowed (platform default) but not excessive here.
Assessment
This skill is internally coherent for finding and recommending agent skills. Before enabling or running it: (1) ensure you have npx installed; (2) review any specific CLIs it will run (npx clawhub, npx skills) before using them, since npx downloads code from npm at runtime; (3) back up AGENTS.md before applying the suggested routing line—only add that line with your explicit consent; and (4) when asked to install a candidate, verify the package/repo and confirm you want to run the install commands (the skill itself emphasizes opt-in installs). If you want extra caution, keep the skill read-only (no AGENTS.md change) and ask the agent to present install strings for manual approval.

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

latestvk97fsh54871xkhsv18mmckz19d82q6n1

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
Binsnpx

Comments