sec-10q-analysis

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTION
Full Analysis
  • EXTERNAL_DOWNLOADS (LOW): The skill relies on npx to fetch and run the octagon-mcp server from the npm registry at runtime. This is the standard deployment model for MCP but introduces a dependency on external package availability and integrity.
  • REMOTE_CODE_EXECUTION (LOW): The references/mcp-setup.md file contains instructions to install Homebrew using a piped bash command: /bin/bash -c "$(curl -fsSL ...)". While this is the official Homebrew installation method, piped remote execution is a high-risk pattern. In this context, it is a documented setup step for the user rather than an automated script behavior.
  • INDIRECT PROMPT INJECTION (LOW): The skill is designed to process external financial documents (SEC 10-Q filings) which could theoretically contain malicious instructions.
  • Ingestion points: SEC 10-Q filings retrieved via the octagon-agent tool.
  • Boundary markers: None explicitly defined in the prompts to distinguish filing content from instructions.
  • Capability inventory: Includes fetching financials and performing deep research.
  • Sanitization: No explicit sanitization or filtering of filing text is mentioned.
  • CREDENTIALS_UNSAFE (SAFE): The documentation correctly uses placeholders (e.g., <your-api-key>, YOUR_API_KEY_HERE) for the required OCTAGON_API_KEY, following security best practices for API configuration instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:32 PM