walrus-ts-sdk
Pass
Audited by Gen Agent Trust Hub on Oct 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates the retrieval of arbitrary data (blobs) from the Walrus network, creating a potential surface for indirect prompt injection where malicious instructions could be embedded in the stored content.
- Ingestion points: Data is ingested via the
client.walrus.readBlobmethod described inSKILL.md. - Boundary markers: There are no specific delimiters or instruction-ignore wrappers enforced for the content of retrieved blobs.
- Capability inventory: The skill enables network operations (interacting with Sui full nodes and Walrus aggregators/publishers), transaction signing, and blob management.
- Sanitization: The skill provides remediation guidance by recommending the use of
@mysten/sealto encrypt sensitive data, but does not sanitize the content of retrieved blobs for potential prompt injection. - [EXTERNAL_DOWNLOADS]: The documentation references several community-maintained SDKs for other languages (Go, PHP, Python) hosted on GitHub. These are provided as informational references for developers working in those environments.
Audit Metadata