walrus-quilts
Pass
Audited by Gen Agent Trust Hub on Oct 14, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides instructions for the agent to execute the 'walrus' CLI tool and 'curl' for storage and retrieval operations. This includes recursive directory traversal to store multiple files and writing downloaded blob content to the local filesystem.
- [INDIRECT_PROMPT_INJECTION]: The skill presents an attack surface for indirect prompt injection by processing untrusted data from local files and remote storage responses.
- Ingestion points: Processes content from local files and directories specified via the '--paths' and '--blobs' flags, and retrieves data from remote Walrus aggregator endpoints.
- Boundary markers: There are no explicit delimiters or system instructions to ignore embedded commands within the files being processed or the API responses.
- Capability inventory: The skill utilizes shell command execution (walrus CLI) and network operations (curl).
- Sanitization: No sanitization or validation logic is defined for the content ingested from external sources.
Audit Metadata