blockscout-analysis

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.90). The SKILL.md workflow requires the agent to fetch and read public Blockscout/Chainscout API responses (including raw proxied responses via direct_api_call and REST endpoints) and explicitly notes those responses can include untrusted third‑party content such as IPFS/HTTP metadata, meaning the agent will consume potentially adversarial, user‑generated data as part of its analysis.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 1.00). The skill mandates calling the MCP tool unlock_blockchain_analysis at runtime via the Blockscout MCP server (e.g. https://mcp.blockscout.com/mcp or https://mcp.blockscout.com/v1/unlock_blockchain_analysis), and the output provides mandatory rules/instructions the agent must follow—i.e., remote content directly controlling agent behavior.