The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied text ("copy") to generate animation node graphs and mutation commands. The ingestion of untrusted content into a workflow that utilizes SDK and CLI capabilities represents an indirect prompt injection surface.
Ingestion points: User-supplied "copy" or "focal phrases" are used as the primary content for animation treatments described in SKILL.md and references/construction-recipes.md.
Boundary markers: The instructions do not define delimiters or specific "ignore embedded instructions" warnings for the user-provided text.
Capability inventory: The skill utilizes CutAgent CLI command cards and SDK mutation routes for Text+, node, and keyframe operations as specified in SKILL.md.
Sanitization: There is no mention of sanitization, escaping, or validation of the input text before it is interpolated into the design and command generation logic.
[DYNAMIC_EXECUTION]: The skill instructs the agent to generate and execute mutations on node graphs, keyframe data, and .setting operations based on the CutAgent SDK and CLI reference.
Evidence: SKILL.md references "exact SDK types, capabilities, identifiers, and mutation routes" for "live capabilities" and "exact Text+, node, keyframe, and .setting operations."
[COMMAND_EXECUTION]: The skill utilizes the CutAgent CLI tool to perform operations within the video editing environment.
Evidence: SKILL.md references "public CutAgent CLI command cards" as authoritative sources for executing operations.

cutagent-motion-text