baoyu-translate

Pass

Audited by Gen Agent Trust Hub on Jun 26, 2026

Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill processes untrusted content from user-provided URLs and files. It lacks explicit boundary markers or delimiters in the templates used to spawn translation subagents (documented in references/subagent-prompt-template.md). This creates a surface for indirect prompt injection where instructions embedded in the source material could potentially influence the agent's behavior.
  • Ingestion points: External URLs, local files, and inline text (see references/workflow-mechanics.md).
  • Boundary markers: None identified in the prompt templates.
  • Capability inventory: Subprocess execution (bun), file system read/write access for intermediate chunks, and network access for content fetching.
  • Sanitization: No evidence of content sanitization or instruction filtering was found.
  • [EXTERNAL_DOWNLOADS]: The skill fetches article content from arbitrary external URLs provided by the user to perform translations, which is its primary functional purpose.
  • [COMMAND_EXECUTION]: The skill executes local TypeScript scripts (scripts/main.ts and scripts/chunk.ts) using the Bun runtime to handle markdown document chunking and metadata processing.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 26, 2026, 08:18 AM
Security Audit — agent-trust-hub — baoyu-translate