The Agent Skills Directory

[PROMPT_INJECTION]: The skill facilitates the ingestion and conversion of external data (JSON or TOON) for use in LLM context, which presents an attack surface for indirect prompt injection.
Ingestion points: The scripts/toon_convert.py script reads data from external files or standard input.
Boundary markers: The skill instructions in SKILL.md recommend wrapping TOON data in specific code fences (e.g., ```toon) to help the model distinguish data from instructions.
Capability inventory: The skill can execute external conversion tools and provide processed text to the LLM context.
Sanitization: No content-level validation or sanitization is performed on the data being converted to prevent embedded instructions from influencing the model.
[COMMAND_EXECUTION]: The Python wrapper script executes external commands to perform data conversion.
Location: scripts/toon_convert.py
Detail: The script uses subprocess.run with a list-based argument structure and no shell execution, which is a secure method for calling external processes.
[EXTERNAL_DOWNLOADS]: The skill relies on external software packages for its core functionality.
Location: scripts/toon_convert.py, SKILL.md
Detail: The skill uses npx to run the @toon-format/cli package from the npm registry. This package is the official tool for the format described and is associated with the skill's author.

toon-format