Security audit
Security checks for vulnerabilities and agentic risk
This skill is a documentation-only workflow for researching ClawHub skills and building clean alternatives, with no executable payload or hidden installation behavior.
This appears safe to install as documentation, but treat its example ideas as starting points only. If you use it to build API, network, or scheduler skills, add explicit user confirmations, scope limits, safe defaults, and warnings around credentials, production systems, and unattended command execution.
The API testing example encourages features such as auth handling, request chaining, and arbitrary HTTP methods without warning that sensitive data, tokens, or credentials may be transmitted to external services. In practice, users may reuse real credentials or target production systems, leading to unintended disclosure, state changes, or privacy impact.
This markdown file proposes building 'net-tools-modern' with port scanning, bandwidth monitoring, and API-based lookups, but does not include any warning that these actions may probe external systems or transmit system/network data. Under the markdown criteria for missing user warnings, descriptions of behaviors affecting privacy or system integrity should disclose those risks.
The task scheduler example proposes scheduling commands, recurring tasks, and conditional execution without warning about the risks of unattended command execution. Even as documentation, this can normalize automation patterns that, if copied into a real skill, may execute destructive commands repeatedly or under unexpected conditions.
No suspicious patterns detected.
