html-skills-listen

Pass

Audited by Gen Agent Trust Hub on May 18, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes a bash script (scripts/listen.sh) to manage the lifecycle of a background receiver process.
  • [COMMAND_EXECUTION]: The bash script spawns a Node.js process (server.js) using nohup to run as a local HTTP server on an ephemeral port.
  • [COMMAND_EXECUTION]: Uses the Monitor tool to tail the server's log file and filter for specific JSON-RPC notifications, bridging external HTTP submissions into the agent's session.
  • [DATA_EXFILTRATION]: The Node.js server specifically binds to 127.0.0.1 (localhost), ensuring the submission endpoint is not exposed to the public network.
  • [INDIRECT_PROMPT_INJECTION]: The skill represents a surface for indirect prompt injection as it ingests data from external HTML artifacts via HTTP POST requests. However, this is the intended primary purpose of the skill, and the risk is mitigated by the use of ephemeral ports and the agent's own processing of the resulting notifications.
Audit Metadata
Risk Level
SAFE
Analyzed
May 18, 2026, 12:57 PM
Security Audit — agent-trust-hub — html-skills-listen