indexer
Pass
Audited by Gen Agent Trust Hub on Oct 10, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill describes methods for the agent to ingest and process data from external blockchain indexer endpoints.
- Ingestion points: GraphQL API responses and WebSocket subscription streams containing block, transaction, and contract state data.
- Boundary markers: The instructions focus on data retrieval and do not specify output delimiters, assuming the chain data is the intended context.
- Capability inventory: Includes network read operations (fetch, WebSocket) and data parsing (hex decoding, BigInt conversion).
- Sanitization: Uses type-safe deserialization from the @midnight-ntwrk SDK to process raw chain data.
- [EXTERNAL_DOWNLOADS]: References official Midnight Network documentation and repositories for API specifications and implementation guides. It also points to standard package registries for required libraries.
- [COMMAND_EXECUTION]: Documents standard installation commands (npm install) for common WebSocket and GraphQL client packages.
Audit Metadata