The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes curl, jq, and python3 via a bash script to communicate with a local Unix socket. It follows security best practices by passing data through environment variables and command arguments, which prevents shell injection.
[DATA_EXFILTRATION]: The skill can read local files to be used as email content. However, the risk of data exfiltration is mitigated by the allowmaild daemon, which only permits delivery to a static list of whitelisted recipient aliases configured by the system administrator. The ingestion point in scripts/send-email.sh (stdin or file-path argument) lacks boundary markers, but its capability is limited to reading files and writing to a local socket. Sanitization is performed by the daemon through strict JSON schema and plain-text validation rules.
[SAFE]: No obfuscation, hardcoded credentials, persistence mechanisms, or remote code execution patterns were detected. The skill is an administrative utility designed for secure interaction with local infrastructure.

allowmail-send