The Agent Skills Directory

[COMMAND_EXECUTION]: The skill documents usage of the openclaw command-line tool to initiate voice calls and check call status. This is standard functionality for the voice-call plugin.
[INDIRECT_PROMPT_INJECTION]: The skill defines tools that ingest untrusted message data which is subsequently processed for voice output.
Ingestion points: The message parameter in the initiate_call, continue_call, and speak_to_user actions (defined in SKILL.md).
Boundary markers: None present in the tool descriptions to delimit user-provided content.
Capability inventory: Uses the openclaw CLI to interface with external telephony providers (Twilio, Telnyx, Plivo).
Sanitization: No mention of content sanitization or safety filtering for the message payload prior to the call initiation.

voice-call