story-short-write
Pass
Audited by Gen Agent Trust Hub on Aug 20, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the AI agent to execute local shell commands for utility purposes.
- Python is used for character counting:
for PYBIN in python3 python py; do ... "$PYBIN" -c "from pathlib import Path; print(len(Path('file').read_text(...)))". - Local Node.js scripts (
check-ai-patterns.js,normalize-punctuation.js,check-degeneration.js) are used to perform style checks, degeneration detection, and punctuation normalization on written content. - These commands are directed at the story's own text files within the workspace.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external reference directories, which constitutes a potential attack surface.
- Ingestion points: The agent reads from
拆文库/(analysis library) and对标/(benchmark files) as mentioned inSKILL.md. - Boundary markers: No specific delimiters or warnings to ignore embedded instructions in these reference files are provided.
- Capability inventory: The agent has the capability to run Python and Node.js scripts as part of its normal workflow.
- Sanitization: There is no specific logic to sanitize or filter potential instructions found in the reference material before processing.
Audit Metadata