zap-fast

Pass

Audited by Gen Agent Trust Hub on Apr 6, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses the cast utility (Foundry toolkit) to perform blockchain operations like balance checks and transaction broadcasting.
  • The script execute-zap.sh performs rigorous validation on all user-supplied inputs using regular expressions (e.g., amount format, hex address patterns, tick ranges) before they are used in shell commands, effectively preventing command injection.
  • A mandatory safety gate requires the agent to obtain a manual "yes" from the user before executing the transaction.
  • A built-in safety threshold blocks transactions exceeding $1,000 USD equivalent unless overridden by the operator.
  • [EXTERNAL_DOWNLOADS]: The skill fetches token information, pricing, and transaction routes from official KyberSwap API endpoints (zap-api.kyberswap.com, token-api.kyberswap.com, aggregator-api.kyberswap.com).
  • These are trusted vendor resources belonging to the skill author (KyberNetwork).
  • [PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection as it processes external token symbols and pool metadata.
  • Ingestion points: User-provided token symbols, pool addresses, and tick ranges; pool details from earn-service.kyberswap.com.
  • Boundary markers: The agent is explicitly instructed to require a verbal confirmation step and to only use this skill when immediate execution is unambiguously requested.
  • Capability inventory: Transaction broadcasting, file system access (for keystores), and network operations.
  • Sanitization: All external data and user inputs are passed through strict regex validation filters in the underlying bash scripts before processing.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 6, 2026, 04:45 AM