The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes several local shell scripts (rmbc-update-check, rmbc-analytics, rmbc-workspace) to manage environment state and telemetry. These are stored within the skill's local directory structure and represent standard functional components.
[DATA_EXFILTRATION]: The skill logs usage statistics via a local script. This telemetry is disclosed to the user and requires an explicit, one-time opt-in via an interactive prompt before any data is processed.
[EXTERNAL_DOWNLOADS]: The skill provides a link to a YouTube video for user training and performs version checks against the vendor's repository. These are typical behaviors for development-oriented tools.
[INDIRECT_PROMPT_INJECTION]: The skill ingests user-provided text for product descriptions and audience analysis to generate sales scripts.
Ingestion points: Inputs defined in SKILL.md such as product_description and target_audience.
Boundary markers: Absent; the skill uses direct interpolation for script generation.
Capability inventory: Executes shell commands for workspace management and accesses local configuration files.
Sanitization: Does not explicitly sanitize user input before it is used in the script generation prompt.

vsl-script