The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill documentation includes instructions to install the required CLI tool by downloading a shell script from https://edgespeak.com/install.sh and piping it directly to the system shell (curl -fsSL https://edgespeak.com/install.sh | sh). This method provides no opportunity for the user or agent to inspect the script before execution, allowing for arbitrary code execution from a remote source.
[COMMAND_EXECUTION]: The skill is designed to execute shell commands using the edgespeak-cli for transcription, status checks, and license activation. It also suggests using curl to interact with a local gateway API on 127.0.0.1.
[INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection because it processes untrusted audio and video files to generate text that the agent then summarizes or acts upon.
Ingestion points: External media files provided by the user at runtime (SKILL.md).
Boundary markers: None identified; the skill does not instruct the agent to use delimiters or ignore embedded instructions within the transcribed text.
Capability inventory: The skill can execute shell commands (edgespeak-cli) and write files to the local system (-o flag).
Sanitization: There is no mention of sanitizing, escaping, or filtering the transcript content before the agent processes it.

edgespeak-transcribe