trailsnap-marketing

Pass

Audited by Gen Agent Trust Hub on Nov 1, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests data from local repository files and external project-related websites to inform content production. This creates a surface for indirect instructions to be processed if the source materials are compromised.
  • Ingestion points: Processes README.md, changelog.md, and content from the official website (trailsnap.cn) and demo site (demo.siyuan.ink).
  • Boundary markers: Relies on structural context and fact-checking instructions, but does not implement explicit delimiters for data isolation.
  • Capability inventory: Performs file writing for marketing materials, generates images via internal tools, and triggers external skills for text humanization.
  • Sanitization: No specific sanitization or filtering logic is described for the ingested text.
  • [EXTERNAL_DOWNLOADS]: The skill accesses the project's official domain and GitHub repository to verify current version information and retrieve branding assets.
  • [COMMAND_EXECUTION]: Utilizes the Playwright automation tool to navigate the product's demo site and capture interface screenshots for use in marketing plogs and tutorials.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 1, 2026, 02:39 PM