Security audit
Security checks for vulnerabilities and agentic risk
The skill’s outreach capabilities are clearly disclosed and include explicit approval requirements, with minor caution needed around broad activation terms.
Prefer the plugin installation and leave approval prompts enabled. Connect only the intended account, keep credentials out of chat, and review recipients, content and client account before approving changes. Clarify unrelated requests if generic words activate LinkedCamp. Approved campaigns and reply agents may continue operating afterward.
Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.
not in openclaw.json:
# ~/.openclaw/.env on the Gateway host
LINKEDCAMP_API_KEY=<paste the key here, never into chat>
The instruction to activate whenever users mention standalone terms such as “prospects” or “sequences” can select this skill for unrelated tasks, introducing LinkedCamp-specific instructions and tool workflows without a clear user request. This creates a limited routing-scope vulnerability, but does not itself authorize data disclosure or external actions. The skill's explicit confirmation requirements, credential protections, and account-selection checks substantially reduce the potential impact.
No suspicious patterns detected.