xl1-patterns
Pass
Audited by Gen Agent Trust Hub on Oct 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill documents patterns for ingesting and processing data from external, untrusted sources including the XL1 blockchain and datalakes, creating an attack surface for indirect prompt injection.
- Ingestion points: Data enters the dApp context through
datalakeViewer.get(hashes)(referenced inchain-data-indexing-protocol.md) andviewer.block.blockByNumber(n)(referenced inchain-data-indexing-client.md). - Boundary markers: The skill emphasizes the use of Zod-based validation (e.g.,
isMovePayload,asMovePayload) to enforce structural integrity, but lacks explicit warnings regarding natural language instructions embedded within string fields. - Capability inventory: The provided patterns include the ability to sign and broadcast transactions to the blockchain using
addPayloadsToChain(referenced inbrowser-ux.mdandcommit-reveal.md). - Sanitization: Employs schema-level validation via Zod factories, which provides structural protection but does not filter for semantic instructions in processed text data.
Audit Metadata