The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes dynamic context injection (using the !command syntax) in SKILL.md to execute cat commands on local project files and shared protocols such as .forgewright/settings.md and skills/_shared/protocols/ux-protocol.md. These commands are used to automatically populate the skill's context with relevant project data and organizational standards at load time.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface because it interpolates the content of external files into its prompt context without using boundary markers or sanitization.
Ingestion points: Content is ingested from files including .forgewright/codebase-context.md, .forgewright/settings.md, and multiple shared protocol files in skills/_shared/ (identified in SKILL.md).
Boundary markers: The skill lacks delimiters or specific instructions to the agent to ignore any malicious commands that might be embedded within the ingested files.
Capability inventory: The skill is designed to autonomously implement complex animation systems, including creating and modifying source code in the src/ directory across multiple phases.
Sanitization: No validation or filtering logic is present to sanitize the content read from these local files before it is processed by the agent.

animation-engineer