arkiv-query
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill defines secure interaction patterns for the Arkiv network, utilizing official vendor SDKs and RPC endpoints owned by the author (arkiv-network).
- [SAFE]: The instructions include a specific security guardrail (Category 8 mitigation) to treat external entity data as untrusted payloads, effectively reducing risks associated with indirect prompt injection from database content.
- [SAFE]: Code examples demonstrate secure credential management, including a custom Python redirect handler (NoRedirect) to ensure that the optional ARKIV_ACCESS_KEY is only transmitted to the fixed official RPC endpoint.
- [SAFE]: The skill implements robust application-level controls such as page and restart budgets for cursors, which prevents potential resource exhaustion or infinite loops during data retrieval.
- [SAFE]: All external references target official documentation, well-known CDNs (unpkg), or vendor-specific infrastructure, with no evidence of typosquatting or malicious obfuscation.
Audit Metadata