The Agent Skills Directory

[SAFE]: The skill does not contain any malicious patterns such as prompt injection, credential theft, or unauthorized data access. Its behavior is entirely consistent with its purpose of generating creative media assets within a controlled environment.
[NO_CODE]: The skill consists exclusively of markdown-based instructions and metadata. It does not include any executable scripts (Python, JavaScript, etc.) or binary files, which inherently limits the technical attack surface.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted user input, such as narration lines or story beats, to generate visual metaphors. While this creates a surface for indirect prompt injection, the risk is mitigated by the skill's specific domain (image/video generation) and its structured workflow. 1. Ingestion points: User-supplied source copy or core concepts as defined in Step 1 of SKILL.md. 2. Boundary markers: The skill does not define explicit delimiters for user input but uses a rigorous style signature to constrain outputs. 3. Capability inventory: Access to platform-native image, video, and audio generation tools. 4. Sanitization: No automated sanitization is described, but the skill enforces two 'Gate' approval steps where a human user must review and confirm the production plan and still frames before video generation occurs, effectively providing a manual sanitization layer.

paper-collage-explainer-generator