The Agent Skills Directory

[DYNAMIC_CONTEXT_INJECTION]: The skill utilizes the bang-backtick syntax to dynamically load content from local documentation files during initialization. Specifically, it executes cat commands to read creator-anti-patterns.md and enforcement-checklist.md from the project's reference directory. This is considered a benign and legitimate use of project-specific tooling to provide the agent with necessary audit context.
[INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied plugin specifications, which creates a potential surface for indirect prompt injection. 1. Ingestion points: User requests for plugin creation, scaffolding, and auditing as defined in the skill metadata. 2. Boundary markers: The instructions do not include specific delimiters or warnings to separate user-provided data from system instructions. 3. Capability inventory: The skill is capable of invoking the plugin-dev:create-plugin tool and modifying plugin manifest files and hooks. 4. Sanitization: The workflow does not specify any sanitization, validation, or filtering of the user's plugin-related input.

plugin-creator