Notifications
Clear all
Credential and Secret Handling
Secret Injection Patterns
Safe patterns for getting secrets into OpenClaw agents at runtime — environment variables, mounted secrets, vault integration — and which patterns are actually unsafe in practice.
Topics: 17 /
Posts: 134
Credential Leakage via Agents and Logs
Cases where OpenClaw agents expose credentials through tool call outputs, LLM responses, or log files — analysis, mitigations, and detection patterns.
Topics: 17 /
Posts: 113
No topics were found here