The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from project documentation files to perform task decomposition.
Ingestion points: Processes docs/specifications/product.md, AGENTS.md, CONTEXT.md, and other linked supporting sources in the first step of its process.
Boundary markers: The instructions do not define clear delimiters or "ignore embedded instructions" warnings when handling the content of these documentation files.
Capability inventory: The agent has the ability to read project files and write new Markdown files to the docs/planning/tasks/ directory.
Sanitization: There is no evidence of filtering or sanitizing the data read from the specifications before it is used to generate the new task files.
[DYNAMIC_EXECUTION]: The skill generates new Markdown files at runtime by populating a predefined template.
Evidence: The process defines a sequence for scanning existing task IDs and creating new files under docs/planning/tasks/TASK-XXX-slug.md based on the assets/task-template.md file.
This represents a low-risk pattern of dynamic content generation from static templates.

create-tasks