skills/1shot-api/skills/1shot-wallet/Gen Agent Trust Hub

1shot-wallet

Pass

Audited by Gen Agent Trust Hub on Aug 28, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill uses official vendor-distributed packages including @1shotapi/ows-provider and @1shotapi/ows-types for the wallet integration.
  • [SAFE]: All external network references and repository links are pointed toward the author's verified infrastructure at wallet.1shotapi.com and the 1Shot-API GitHub organization.
  • [INDIRECT_PROMPT_INJECTION]: The skill describes a surface for ingesting external data through host configurations and blockchain metadata.
  • Ingestion points: Untrusted data enters the agent context via configuration options in the configure RPC and metadata retrieved from the blockchain during asset discovery.
  • Boundary markers: The skill documentation specifies that the underlying library uses Zod strict schema validation to enforce expected data structures.
  • Capability inventory: The skill enables transaction signing via EIP-1193 and network communication for analytics and relay services.
  • Sanitization: Input is validated using strict schemas to filter out unexpected properties or injection attempts within configuration objects.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 28, 2026, 04:44 AM
Security Audit — agent-trust-hub — 1shot-wallet