monetize-service
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructions require the installation of third-party Node.js packages from the npm registry, including 'express' and 'x402-express'. While 'express' is a standard library, 'x402-express' is a less common dependency specific to this skill's functionality.
- [REMOTE_CODE_EXECUTION]: The skill utilizes 'npx awal@latest', which fetches and executes the latest version of the 'awal' package from the npm registry at runtime. This pattern involves running remote code that is not version-pinned, although it is central to the skill's intended workflow for wallet management and payment testing.
- [INDIRECT_PROMPT_INJECTION]: The skill provides example code for an Express server that ingests external data via 'req.body' (e.g., in the sentiment analysis endpoint). This creates a vulnerability surface where the resulting agent-powered service could process malicious instructions embedded in user-supplied data. Evidence Chain: 1. Ingestion points: 'req.body' in the POST /api/analyze example within SKILL.md. 2. Boundary markers: Absent in the provided code snippets. 3. Capability inventory: The resulting server can process requests and return JSON responses. 4. Sanitization: The boilerplate code does not include explicit validation or sanitization for the 'text' input field.
Audit Metadata