The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a wide range of system commands for infrastructure probing, including kubectl, docker, colima, and launchctl.
[COMMAND_EXECUTION]: The skill allows for the definition and execution of arbitrary shell commands via the [script.<name>] section in the services.toml configuration file, which are executed using sh -c.
[REMOTE_CODE_EXECUTION]: As noted in the troubleshooting instructions and automated scans, the skill documentation suggests piping the output of a network-accessible health endpoint (http://127.0.0.1:9999/health) directly into a Python interpreter for JSON processing.
[COMMAND_EXECUTION]: The skill performs remote command execution via SSH to monitor the internal state of virtual machines (e.g., lima-colima), including executing Python-based socket probes.
[EXTERNAL_DOWNLOADS]: The skill interacts with external APIs, specifically the Telegram Bot API, to send SOS notifications and alerts during infrastructure failures.
[DATA_EXFILTRATION]: System status data and critical error logs are transmitted to external communication channels (Telegram and iMessage) as part of the escalation process.
[COMMAND_EXECUTION]: The skill manages system-level persistence and service lifecycles through launchctl, including the ability to 'kickstart' or 'bootout' services.

talon