story-long-write
Pass
Audited by Gen Agent Trust Hub on Aug 20, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill invokes local utility scripts within its package to manage the novel's state and verify prose quality. Specifically, it uses a Python script (
scripts/tracking_commit.py) to synchronize story data and multiple Node.js scripts (scripts/check-ai-patterns.js,scripts/check-degeneration.js,scripts/normalize-punctuation.js) to perform automated linting and formatting on the generated text. These executions are internal to the skill's workflow. - [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by reading data from novel benchmark libraries and project-specific files (e.g.,
选题决策.md). This allows for the ingestion of untrusted data into the agent's context. However, the skill includes defensive design patterns such as automated prose validation scripts and specialized review agents (Phase 5) to ensure the agent maintains narrative control and quality.
Audit Metadata