walrus-overview

Pass

Audited by Gen Agent Trust Hub on Jul 6, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is strictly documentation-based and serves as an entry point for users to understand Walrus architecture and terminology.
  • [PROMPT_INJECTION]: No attempts to override agent behavior or bypass safety guidelines were found. The 'Rules' section contains benign operational constraints.
  • [DATA_EXFILTRATION]: There are no network tools or commands that access sensitive files. The skill correctly identifies that data stored on Walrus is public by default and recommends encryption for privacy.
  • [REMOTE_CODE_EXECUTION]: The skill does not contain any code snippets that download or execute remote scripts. It mentions external tools (CLI, SDK) for user reference only.
  • [COMMAND_EXECUTION]: No shell commands or dynamic execution patterns are present in the skill files.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 6, 2026, 07:24 AM
Security Audit — agent-trust-hub — walrus-overview