The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The skill manages highly sensitive files, including the generation of private ed25519 keys and the modification of the system-wide SSH configuration file (sshd_config), which contains security policies and authentication settings.
[PERSISTENCE]: The workflow involves adding public keys to the ~/.ssh/authorized_keys file and configuring persistent SSH tunnels and aliases in ~/.ssh/config. These operations establish long-term access mechanisms to the target system.
[COMMAND_EXECUTION]: The skill relies on the agent executing various system binaries such as ssh-keygen, ssh-copy-id, and the ssh client for creating tunnels and managing remote identities.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to read and process system logs via the homelab_serviceLogs tool to diagnose connection issues, which introduces an attack surface where malicious log entries could influence agent behavior. * Ingestion points: System logs are read from the Raspberry Pi's SSH daemon via tool output. * Boundary markers: No specific delimiters or instructions are provided to the agent to distinguish between log data and core instructions. * Capability inventory: The agent possesses capabilities to perform file writes (authorized_keys), modify system configurations (sshd_config), and restart system services. * Sanitization: There is no evidence of sanitization, filtering, or validation applied to the log content before it is processed by the agent.

ssh-management