The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user-provided content (markdown files) and possesses file-writing capabilities through the Write and Edit tools.
Ingestion points: The skill ingests data from local .md files specified by the user or discovered in directories via Glob, Grep, and Read tools (SKILL.md, README.md).
Boundary markers: The skill does not explicitly use instruction-ignoring delimiters or boundary markers when the agent reads and analyzes the input files, although it includes a rule (S.1) to apply XML tags to the generated output.
Capability inventory: The skill is granted access to the Read, Write, Edit, Grep, and Glob tools, enabling it to perform extensive filesystem operations.
Sanitization: The skill does not specify any sanitization, validation, or filtering of the external content before it is processed by the agent's logic.

text-optimizer