The Agent Skills Directory

[COMMAND_EXECUTION]: The skill facilitates the execution of the edgespeak-cli tool. It instructs the agent to run various subcommands such as status, login, activate, transcribe, and segment. These commands are specific to the EdgeSpeak vendor's transcription utility and are used to manage local transcription workflows and licensing.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data in the form of audio and video files provided by the user.
Ingestion points: Media files are passed to edgespeak-cli transcribe for processing into text or JSON.
Boundary markers: The instructions do not specify the use of delimiters or 'ignore' instructions when handling the resulting transcript text.
Capability inventory: The skill has the capability to execute CLI commands, write files to disk (using the -o flag), and make network requests to a local gateway (127.0.0.1).
Sanitization: There is no mention of sanitizing or escaping the transcription output before the agent is asked to 'summarize / clean up / translate' the content. This creates a surface where malicious instructions embedded in the audio could potentially influence the agent's behavior during these post-processing steps.

edgespeak-transcribe