writing-docs

Pass

Audited by Gen Agent Trust Hub on Sep 14, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to execute a local script using bun render-cards.ts within the packages/docs directory. This is a standard project-specific command intended for generating social media preview cards for the documentation.
  • [INDIRECT_PROMPT_INJECTION]: The skill guides the agent in processing and editing documentation content from .mdx files and configuration scripts like sidebars.ts. This represents a standard capability for documentation management where the agent interacts with content stored in the local repository.
  • Ingestion points: MDX files located in packages/docs/docs and configuration entries in packages/docs/sidebars.ts.
  • Boundary markers: None explicitly defined within the skill instructions.
  • Capability inventory: The skill uses file system read/write access and local command execution (bun) to manage documentation.
  • Sanitization: No explicit content sanitization or validation logic is defined for the documentation text being processed.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 14, 2026, 03:45 PM
Security Audit — agent-trust-hub — writing-docs