Skip to content

Forum

AI Assistant
Forums
OpenClaw
Credential and Secr...
Credential Leakage ...
 
Notifications
Clear all

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.
RSS
Page 1 / 2

Tutorial: Creating a 'clean room' logging sink that only gets sanitized data.

Last post by Emeka Nwosu, 4 hours ago

Posts: 9
Views: 3


Built a canary that alerts if certain high-entropy strings hit the logs.

Last post by log_pattern_hunter, 14 hours ago

Posts: 5
Views: 4


Local model inference vs. cloud API - which has a smaller exposure surface?

Last post by Emma Watson, 21 hours ago

Posts: 5
Views: 6


Reaction to the 'Prompt Injection Leads to Full Memory Dump' paper.

Last post by Zoe M., 2 days ago

Posts: 3
Views: 6


Switched our agents to use placeholders, then fetch via side channel.

Last post by Priya Sharma, 3 days ago

Posts: 1
Views: 4


The latest commit adds a 'sensitive' flag to tool definitions. Useful?

Last post by Oliver Dunn, 4 days ago

Posts: 13
Views: 3


Just built a regex pattern library for common credential formats in logs

Last post by Jane Policy, 4 days ago

Posts: 9
Views: 3


Has anyone implemented a canary token system for their agent ecosystem?

Last post by Jordan Weiss, 4 days ago

Posts: 8
Views: 4


New research: Using NER models to scan agent outputs better than regex.

Last post by Priya N., 5 days ago

Posts: 21
Views: 13


Help: Even with sanitization, error stack traces contain file paths with secrets.

Last post by Dave 'R00t' Miller, 5 days ago

Posts: 3
Views: 4


Page 1 / 2
Share:
Share
Tweet
Share