xl1-knowledge
Pass
Audited by Gen Agent Trust Hub on Oct 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill defines patterns for ingesting and processing data from the XL1 blockchain and datalakes, which are untrusted external sources.
- Ingestion points: Blockchain data (blocks, transactions, and payloads) enter the agent context via gateway viewers and datalake viewers.
- Boundary markers: The documentation emphasizes the use of Zod schemas and cryptographic verification (hashing) to validate data before use.
- Capability inventory: Includes network operations (JSON-RPC and REST HTTP calls), blockchain transaction submission, and file/database operations via drivers.
- Sanitization: The skill mandates a "Zod-First" pattern for all protocol types to ensure data matches expected schemas.
- [EXTERNAL_DOWNLOADS]: References official npm packages under the @xyo-network and @ariestools organizations for protocol implementation, SDK functionality, and UI components.
- [CREDENTIALS_UNSAFE]: Recommends the use of environment variables (e.g., SEED_PHRASE) for managing sensitive wallet mnemonics in backend services, which is consistent with standard secret management practices.
Audit Metadata