The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes a local script scripts/check_slop.py and a test runner tests/run_tests.py. These scripts are intended to be executed by the agent to perform deterministic validation of writing drafts against the style guide's rules. This is a functional component of the skill's self-correction loop.
[EXTERNAL_DOWNLOADS]: The skill directs the agent to fetch the official Apple Style Guide from support.apple.com to ensure technical documentation adheres to industry standards. This reference targets a well-known and trusted service.
[INDIRECT_PROMPT_INJECTION]: The skill processes user-provided text for rewriting and auditing, which introduces a potential surface for indirect prompt injection. However, the risk is mitigated by the skill's strict stylistic constraints and the automated validation process.
Ingestion points: User-supplied text provided for writing, drafting, or auditing tasks (identified in SKILL.md).
Boundary markers: None identified; the skill does not explicitly instruct the agent to use delimiters when ingesting user content.
Capability inventory: The agent is authorized to write drafts to local files, execute a local Python validation script (scripts/check_slop.py), and access the external Apple Style Guide website.
Sanitization: The skill provides a mechanical check script (scripts/check_slop.py) that flags stylistic violations, although it does not perform explicit security sanitization of input instructions.

deslop-writing