security-sentinel

You can scan the workspace for exposed secrets by running predefined patterns that detect API keys, private keys, and tokens across various file types to prevent accidental credential exposure.

Auditing agent configurations involves checking for critical missing files like SECURITY.md and AGENTS.md, and detecting suspicious or unauthorized directories to ensure adherence to security best practices.

You need continuous security monitoring in autonomous agent environments, especially before deploying a new agent or after significant code changes, to perform risk assessment and prevent system compromises.

Yes, this security scanner operates without external dependencies, relying on internal scripts to perform file integrity checks and pattern matching for vulnerability detection and configuration auditing.

File integrity checks and predefined patterns are used to detect security vulnerabilities including exposed secrets, misconfigurations, and suspicious or unauthorized forbidden paths within the agent's workspace.

Yes, you can identify suspicious or unauthorized directories through forbidden path detection, which applies predefined patterns to scan the current workspace for potential security risks and exposure.