chapter-writing

Pass

Audited by Gen Agent Trust Hub on Nov 11, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes a local command-line tool named story and provides a fallback using node to run a relative script (../story-maintenance/scripts/story.js). These commands are used for project maintenance tasks like reindexing, word counts, and continuity checks.
  • [INDIRECT_PROMPT_INJECTION]: The skill identifies the risk of processing untrusted data (like reader notes or editor comments) and explicitly instructs the agent to treat such content as reference material rather than instructions. This follows security best practices for handling external data. The analysis identifies ingestion points in the story project files and user notes, with a clear boundary marker instructing the agent to treat external text as data only, while capabilities are limited to project-specific CLI calls and file writes.
  • [DYNAMIC_EXECUTION]: The instructions specifically prohibit the use of project-local generator or build scripts (e.g., build-*.js) for creating chapters, which prevents the accidental execution of potentially malicious code contained within a user's story project files.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 11, 2026, 10:26 AM