The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill provides a framework for processing external data and user-provided syllabuses or books to generate curriculum content, creating a vulnerability surface for indirect prompt injection.
Ingestion points: Untrusted data enters the agent's context through the fetch_url tool (used to read books or external sources) and via detailed user responses gathered through the ask_user tool.
Boundary markers: The skill instructions do not specify the use of delimiters or clear directives to ignore instructions that might be hidden within the external content or user-supplied reference materials.
Capability inventory: The skill possesses significant capabilities, including file system operations (create_folder, create_stage, patch_stage), content generation (generate_scene, generate_tts), and network access (web_search, fetch_url).
Sanitization: There are no explicit instructions for the agent to sanitize, validate, or filter content retrieved from external sources before it is interpolated into prompts for generating new lesson content.

curriculum-planner