The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes the pipecat CLI tool to list options and scaffold projects. These commands are constructed based on user-selected options and are part of the tool's primary purpose. Safety is maintained by requiring user confirmation before execution and relying on a pre-installed CLI tool (pipecat).
[INDIRECT_PROMPT_INJECTION]: The skill ingests data from the output of pipecat init --list-options to dynamically generate user prompts. While this is an ingestion point for external data, the data source is a local CLI tool, and the risk is minimal as the input is used only to populate selection lists for the user. Mitigation includes providing the user with a final summary of all choices for verification before the project is scaffolded.

pipecat-init