trader-portfolio-cg

Pass

Audited by Gen Agent Trust Hub on May 20, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill installs the neural-trader package via npm. It explicitly uses the --ignore-scripts flag, which is a critical security measure to prevent the execution of arbitrary or malicious lifecycle scripts during the installation process.\n- [COMMAND_EXECUTION]: Executes npx neural-trader to retrieve current portfolio metrics and as a fallback for optimization. These operations are limited to the specific utility required for the skill's financial calculations and no unauthorized commands were found.\n- [DATA_EXFILTRATION]: No external data exfiltration was detected. Data transmission is limited to local command outputs and designated MCP memory namespaces for storage and cross-checking as defined in the trading-risk namespace.\n- [PROMPT_INJECTION]: The skill identifies an indirect prompt injection surface because it processes data retrieved from external tools.\n
  • Ingestion points: Data is read from the neural-trader CLI JSON output and via the mcp__claude-flow__memory_search tool (SKILL.md).\n
  • Boundary markers: The instructions do not define specific delimiters or instructions to the agent to disregard potential instructions embedded within the ingested data.\n
  • Capability inventory: The skill possesses shell execution capabilities (Bash) and access to persistent memory storage tools (mcp__claude-flow__memory_store) as seen in SKILL.md.\n
  • Sanitization: The instructions do not specify any validation or sanitization of the input data before it is utilized in the optimization workflow.
Audit Metadata
Risk Level
SAFE
Analyzed
May 20, 2026, 04:39 AM
Security Audit — agent-trust-hub — trader-portfolio-cg