The Agent Skills Directory

[SAFE]: The skill facilitates interaction with the author's official infrastructure at api.varg.ai and mcp.varg.ai. All network operations are directed towards these vendor-controlled endpoints.
[SAFE]: Secret management follows industry best practices by recommending the use of environment variables ($VARG_API_KEY) for API authentication rather than hardcoding sensitive credentials.
[SAFE]: The use of command-line tools such as curl, jq, and ffmpeg is limited to standard, legitimate tasks like making API requests to the generative service and trimming output media files.
[INDIRECT_PROMPT_INJECTION]: The skill defines a workflow where user-supplied prompts and image URLs are interpolated into shell commands.
Ingestion points: User-provided one-line prompts and image URLs in SKILL.md (Section 4).
Boundary markers: None explicitly defined in the provided shell command examples.
Capability inventory: The skill utilizes Bash(curl), Bash(jq), and Bash(ffmpeg) and has access to file system tools (Read, Write, Edit).
Sanitization: No explicit input validation or escaping is described in the documentation, relying on the agent's standard safety protocols and tool execution environment.

kling-4-action-montage