skills/sendaifun/skills/kamino/Gen Agent Trust Hub

kamino

Pass

Audited by Gen Agent Trust Hub on Sep 14, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill provides templates for building agents that ingest external data from Scope oracles and Solana blockchain state to automate financial decisions. This ingestion of untrusted external feeds introduces a surface for indirect prompt injection where manipulated external data could influence the agent's logic for liquidations or rebalances.
  • Ingestion points: Real-time price data and protocol configurations are fetched via @kamino-finance/scope-sdk and @kamino-finance/klend-sdk in files such as templates/full-integration.ts and examples/oracle/price-feeds.md.
  • Boundary markers: The skill encourages logical boundaries such as health factor monitoring (e.g., alertThreshold = 1.2) and slippage protection (e.g., defaultSlippageBps = 50).
  • Capability inventory: The skill enables the agent to execute financial transactions involving user funds, including deposits, borrows, and strategy management using standard Solana transaction methods.
  • Sanitization: Mathematical operations are handled by decimal.js for precision, and example code includes checks for price staleness (maxStaleSlots) and confidence ratios.
  • [EXTERNAL_DOWNLOADS]: The skill documentation and examples reference official Kamino Finance SDKs and Solana development libraries for installation via NPM.
  • Packages: @kamino-finance/klend-sdk, @kamino-finance/kliquidity-sdk, @kamino-finance/scope-sdk, @solana/web3.js, @coral-xyz/anchor, decimal.js.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 14, 2026, 10:06 PM