The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to use shell commands such as grep, tail, awk, and docker exec to perform system-level log analysis and process monitoring. While these are appropriate for the intended use case, they grant the agent significant control over the MikoPBX container environment.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze log data from various sources (nginx, PHP, Asterisk) that reflect external activity. This creates an attack surface where an attacker could trigger specific log entries containing malicious instructions to influence the agent's behavior.
Ingestion points: Multiple log files including system/messages, php/error.log, and nginx/access.log that record untrusted data from network interactions.
Boundary markers: The skill lacks instructions for the agent to use markers or delimiters to separate its instructions from the content of the logs being analyzed.
Capability inventory: The agent has access to Bash tools, docker exec, and utility commands like ps and redis-cli which could be misused if an indirect injection is successful.
Sanitization: There are no explicit requirements for the agent to sanitize or validate the log content before processing.

log-analyzer