sb-stories

Pass

Audited by Gen Agent Trust Hub on Oct 2, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill analyzes local source code (e.g., React components) to detect state branches and property shapes. The results are used to drive agent behavior, creating a potential surface where malicious code structures or comments in processed files could influence the agent's subsequent actions.
  • [DYNAMIC_EXECUTION]: Several scripts automate the generation and modification of project files. scripts/scaffold-page-story.py generates React story files from templates, and scripts/scaffold-factory.sh appends boilerplate factory definitions to project files.
  • [COMMAND_EXECUTION]: The skill uses shell utilities and subprocess calls to analyze the project filesystem. Scripts like extract-states.sh and extract-prop-shapes.sh use grep, awk, and sed to scan source files, while scaffold-page-story.py invokes sibling Python scripts to process metadata.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 2, 2026, 09:23 AM