baoyu-post-to-x

Fail

Audited by Gen Agent Trust Hub on Aug 14, 2026

Risk Level: HIGHEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Fetches the Bun runtime installer from bun.sh, a well-known service, and pipes it to a shell for execution as part of the environment setup instructions.
  • [EXTERNAL_DOWNLOADS]: The md-to-html.ts script downloads remote images from arbitrary URLs specified within user-provided Markdown files to a local temporary directory.
  • [COMMAND_EXECUTION]: The skill extensively uses system-level commands including osascript (macOS), powershell (Windows), and xdotool/ydotool (Linux) to simulate user keyboard input (e.g., Cmd+V) to bypass browser-level automation detection.
  • [COMMAND_EXECUTION]: Instructs the agent to use pkill to terminate existing Chrome processes with active debugging ports to resolve connectivity conflicts.
  • [REMOTE_CODE_EXECUTION]: Dynamically generates Swift scripts for macOS clipboard management, which are written to temporary files and executed via spawnSync.
  • [PROMPT_INJECTION]: The skill processes external Markdown files to generate HTML articles, creating a surface for indirect prompt injection.
  • Ingestion points: markdownPath parameter in scripts/x-article.ts and scripts/md-to-html.ts.
  • Boundary markers: Absent; the skill does not use explicit delimiters or instructions to ignore embedded commands within the source data.
  • Capability inventory: Includes spawnSync for shell commands, osascript for input simulation, pkill for process management, and Chrome DevTools Protocol (CDP) for browser control.
  • Sanitization: Performs standard character escaping and Markdown parsing using the marked library.
Recommendations
  • HIGH: Downloads and executes remote code from: https://bun.sh/install - DO NOT USE without thorough review
Audit Metadata
Risk Level
HIGH
Analyzed
Aug 14, 2026, 04:00 PM
Security Audit — agent-trust-hub — baoyu-post-to-x