The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user data, specifically the 'vibe' and 'source basis', and interpolates these strings directly into the --prompt parameter of a shell command (postplus media create). This exposes the agent to indirect prompt injection where a user could attempt to manipulate the downstream image generation tool or the agent's behavior through the manifest drafting process.
Ingestion points: User-provided vibe, platform, and source basis described in SKILL.md.
Boundary markers: No explicit delimiters or instructions to ignore embedded instructions are present when generating prompts or handling the manifest.
Capability inventory: The skill executes postplus CLI tools for media creation, schema management, and quote confirmation (SKILL.md).
Sanitization: There is no evidence of input validation or escaping for the prompt string before it is passed to the shell command.
[COMMAND_EXECUTION]: The skill utilizes a suite of CLI tools (postplus) to manage slideshow manifests, query media schemas, and orchestrate batch image generation. These commands represent the core functionality for producing TikTok and Instagram content but involve direct interaction with the host system's shell environment.

slideshow-producer