cultural-tourism-xiaohongshu-feed

Fail

Audited by Gen Agent Trust Hub on Aug 29, 2026

Risk Level: HIGHPERSISTENCECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [PERSISTENCE]: The skill implements a subscription feature that creates a persistent daily task on the user's system. On macOS, it generates and loads a LaunchAgent plist file (~/Library/LaunchAgents/com.qoder.cultural-tourism-xiaohongshu-feed.plist). On Linux, it modifies the user's crontab to schedule daily execution of the Python script.- [COMMAND_EXECUTION]: The script uses the subprocess module to execute shell commands. This includes calling 'launchctl' to load and unload background services, 'crontab' to modify scheduled tasks, and 'open' or 'xdg-open' to launch the system's web browser for displaying the generated HTML reports.- [INDIRECT_PROMPT_INJECTION]: The skill fetches social media notes from an external API (Xiaohongshu) and processes their titles and metadata. This content is then used to generate summaries and terminal tables. There is an attack surface where malicious instructions embedded in the external post titles could attempt to influence the agent's output or the user's browser environment.
  • Ingestion points: External content is ingested via the 'fetch_articles' function in 'scripts/cultural_tourism_xiaohongshu_report.py' which queries a remote API.
  • Boundary markers (absent): The skill lacks explicit boundary markers or instructions for the agent to ignore potential commands within the fetched social media data.
  • Capability inventory: The skill possesses capabilities for network requests (requests.post), file writing (reports and plists), and shell command execution (subprocess) within 'scripts/cultural_tourism_xiaohongshu_report.py'.
  • Sanitization (absent): The 'sanitize_title' function provides minimal sanitization by replacing brackets and newlines to avoid markdown formatting issues, but it does not filter for potential prompt injection payloads.- [EXTERNAL_DOWNLOADS]: The script performs network operations to fetch data from 'https://redfox.hk/story/api/parseWork/queryXhsPlayletMsgs'. This is identified as a vendor-owned resource for data retrieval.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Aug 29, 2026, 02:38 AM
Security Audit — agent-trust-hub — cultural-tourism-xiaohongshu-feed