sui-seal

Pass

Audited by Gen Agent Trust Hub on Sep 23, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill includes instructions to install @mysten/seal and @mysten/sui via npm. These are standard packages for interacting with the Sui blockchain and Mysten Labs' Seal service.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process external data for decryption, which presents a surface for potential indirect instructions.
  • Ingestion points: The SealClient.decrypt and SealClient.encrypt methods in SKILL.md process external encryptedObject data from sources like Walrus or Sui objects.
  • Boundary markers: None are specified for the raw byte data being decrypted.
  • Capability inventory: The skill uses SuiGrpcClient for network operations and provides cryptographic capabilities for local decryption.
  • Sanitization: The skill notes that once data is decrypted, the application is responsible for its handling, emphasizing that Seal has no visibility into the resulting plaintext.
  • [SAFE]: The skill references official documentation (seal-docs.wal.app) and the official Mysten Labs repository on GitHub for pattern implementations, which are appropriate for a technical integration guide.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 23, 2026, 11:21 AM
Security Audit — agent-trust-hub — sui-seal