skills/sendaifun/skills/orca/Gen Agent Trust Hub

orca

Pass

Audited by Gen Agent Trust Hub on Sep 14, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill uses official and well-known SDK packages from the Orca organization. All network interactions are directed toward standard Solana RPC infrastructure (e.g., api.mainnet-beta.solana.com) or reputable RPC providers such as Helius and QuickNode. Code examples demonstrate loading wallet keys from a local 'keypair.json' file, which is a standard development practice for the Solana blockchain ecosystem.
  • [EXTERNAL_DOWNLOADS]: The guide instructs users to install packages from the npm registry, including @orca-so/whirlpools, @solana/kit, and @coral-xyz/anchor. These are recognized libraries within the Solana developer community and originate from well-known services.
  • [INDIRECT_PROMPT_INJECTION]: The skill provides templates that ingest user-provided blockchain data, such as token mint addresses and pool addresses, to perform swaps and manage liquidity positions.
  • Ingestion points: Variable inputs for pool addresses and token amounts in files like 'templates/setup.ts' and 'examples/swap/token-swap.ts'.
  • Boundary markers: Absent; the instructions do not include explicit delimiters or warnings for the agent to ignore potentially malicious instructions embedded in transaction data or pool metadata.
  • Capability inventory: The skill utilizes file system read access for wallet key loading and network operations for interacting with Solana RPC nodes.
  • Sanitization: The examples rely on the Orca SDK for address and transaction validation without demonstrating additional user-level input sanitization.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 14, 2026, 10:06 PM
Security Audit — agent-trust-hub — orca