The Agent Skills Directory

[PROMPT_INJECTION]: The skill includes functionality to process external performance data, which presents an indirect prompt injection surface.
Ingestion points: Performance data from CSVs, clipboard pastes, or API outputs specified in SKILL.md.
Boundary markers: None identified; the instructions do not include specific delimiters to isolate external data from the prompt logic.
Capability inventory: The skill is capable of executing shell commands and network requests using node, npx, and curl.
Sanitization: No explicit sanitization or filtering of the input data is described.
[COMMAND_EXECUTION]: The skill documentation describes a workflow involving the execution of various command-line tools.
Examples include platform-specific CLI scripts (e.g., google-ads.js) and video rendering via npx remotion render.
It also provides curl examples for interacting with external AI APIs.
[EXTERNAL_DOWNLOADS]: The skill references external source code and tools from public repositories and services.
It includes instructions for cloning the Voicebox repository from GitHub and fetching Remotion templates via npx.
These references target well-known development platforms and established services.

ad-creative