global-chat-agent-discovery

Pass

Audited by Gen Agent Trust Hub on Jul 3, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs users to install a remote package from an external registry using npm install -g @global-chat/mcp-server. While this package is associated with the skill's authoring organization (Global Chat), it involves the acquisition of external code.
  • [COMMAND_EXECUTION]: The skill provides instructions to execute shell commands, specifically npm install and npx -y @global-chat/mcp-server, which run code on the host environment to initialize the MCP server functionality.
  • [PROMPT_INJECTION]: The skill presents a surface for indirect prompt injection (Category 8) because it ingests and processes data from an external, community-driven directory of 18,000+ entries.
  • Ingestion points: Data enters the agent's context through search results returned by the @global-chat/mcp-server or the global-chat.io web directory.
  • Boundary markers: The instructions do not define specific delimiters or instructions to ignore embedded prompts within the search results.
  • Capability inventory: The skill encourages the use of npm install and npx to set up tools, and the resulting server enables discovery of other tools the agent might be persuaded to execute.
  • Sanitization: There is no evidence of automated sanitization or filtering of the search result content before it is processed by the AI agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 3, 2026, 01:07 PM
Security Audit — agent-trust-hub — global-chat-agent-discovery