The Agent Skills Directory

[SAFE] (SAFE): The analyzed files are reference materials (markdown and Python snippets) focused on D-Bus patterns and security. There are no executable scripts or malicious instructions.
[INDIRECT_PROMPT_INJECTION] (LOW): The skill provides patterns for processing external D-Bus data. While this presents an attack surface, the included security-examples.md provides specific mitigations including regex-based input validation and service allowlisting.
[DATA_EXPOSURE] (INFO): The Peer Credential Validation snippet reads from /proc/{pid}/exe. This is an intentional security check to verify the identity of the process connecting to the D-Bus, not a data leak.

dbus