The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill's workflow involves reading and processing prompt excerpts from references/cases.md, which creates a surface for indirect prompt injection where instructions embedded in the examples could influence agent behavior. 1. Ingestion points: The agent is instructed to read references/cases.md on every invocation. 2. Boundary markers: The skill lacks explicit delimiters or instructions to ignore embedded commands within the ingested text. 3. Capability inventory: The agent is directed to use browser, image, and video tools to inspect external media links. 4. Sanitization: No sanitization or filtering of the prompt excerpts or tool outputs is performed.
[EXTERNAL_DOWNLOADS]: The skill directs the agent to visit external domains such as goodcase.ai and x.com to gather visual evidence for its creative output. These references are documented as part of the skill's intended evidence-gathering workflow.

epic-fantasy-scifi-spectacle-by-1r2e570