The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses subprocess.run to execute local Python scripts (writing_section_index.py, writing_gate_probe.py, writing_prose_gate.py) and a shared engine (prose-audit.py) located within the plugin directory. These executions are used for mechanical verification of document grammar, citations, and prose style.\n- [DYNAMIC_EXECUTION]: The test suite (scripts/writing_section_index_test.py) utilizes importlib.import_module to dynamically load the writing_section_index module for unit testing purposes. This is a standard testing practice and occurs in a controlled environment.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided documents, creating an attack surface for indirect prompt injection.\n
Ingestion points: Drafts, outlines, and plan files in the project directory.\n
Boundary markers: Markdown headings are used for structure, though explicit instructions to ignore embedded instructions are absent.\n
Capability inventory: Tools for file system modification (Write, Edit) and shell execution (Bash) are available to the agent.\n
Sanitization: Automated scripts mechanically validate plan hashes, claim identifiers, and citation keys to ensure integrity between planned and drafted content.

writing