The Agent Skills Directory

[SAFE]: No security issues were detected during the analysis of the documentation or the parser script.
[COMMAND_EXECUTION]: The scripts/toon_parser.py module uses only Python standard libraries (json, re, dataclasses, typing). It does not use any functions for executing shell commands, spawning subprocesses (e.g., os.system, subprocess.run), or performing dynamic code evaluation (e.g., eval, exec).
[DATA_EXFILTRATION]: No network-related operations or hardcoded credentials were found. The parser operates purely on string data and does not interact with external APIs or the local file system.
[DYNAMIC_EXECUTION]: The parser includes a safe usage of json.loads() to process embedded JSON structures within quoted strings. This is a standard and secure practice for data deserialization in Python.
[PROMPT_INJECTION]: The skill is defined as a 'knowledge' mode reference. It contains no instructions that attempt to override agent safety guidelines or bypass restrictions. The provided agent prompt patterns are educational and follow standard handoff conventions.
[INDIRECT_PROMPT_INJECTION]: While the format is intended for processing data from other agents (an inherent surface for indirect injection), the parser itself does not provide any capabilities that could be exploited by malicious data content. It simply transforms the TOON format into a standard Python dictionary.

ref-toon-format