story-review
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes local Node.js and Python scripts (e.g.,
normalize-punctuation.js,check-ai-patterns.js,author_memory_commit.py) to perform story analysis and manage author preferences. These operations are strictly limited to the project's local workspace and utility tasks. - [DYNAMIC_EXECUTION]: The script
tracking_commit.pydynamically loads a local helper modulewordcount_core.py. This is used for modular code organization within the skill package and does not involve loading untrusted or remote code. - [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided novel text across multiple subagents.
- Ingestion points: Reads novel text from
正文/and大纲/directories. - Boundary markers: Explicitly instructs agents to use a structured
Findings Schemaand uses script-level masking (maskQuoted) to distinguish dialogue from narrative text during checks. - Capability inventory: The skill can read/write files in
追踪/and.story/via helper scripts to maintain state. - Sanitization: Implements automated punctuation normalization and pattern detection to ensure consistent formatting before analysis.
Audit Metadata