The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes novel text and project metadata (YAML files) which could contain hidden instructions, leading to indirect prompt injection.
Ingestion points: Reads novel segments and configuration files like plan.yaml, memory.yaml, and continuity.yaml from the local environment.
Boundary markers: Lacks explicit delimiters or instructions to ignore potential commands embedded within the processed text.
Capability inventory: The skill can modify local markdown and YAML files and execute local Python scripts.
Sanitization: No sanitization of the novel text before processing is specified.
[COMMAND_EXECUTION]: The skill executes local Python scripts for administrative tasks.
Evidence: Uses novel-studio/tools/word_count.py and skills/ns-write/scripts/chapter_audit.py to calculate word counts and audit chapters after writing.

ns-revise