upstash-search-js

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill documentation includes instructions for installing @upstash/search via NPM and importing upstash_search in Python. These are the official libraries provided by the vendor (Upstash) for interacting with their Search service.
  • [INDIRECT_PROMPT_INJECTION]: The skill facilitates the retrieval of documents and metadata from a search index, creating a surface for indirect prompt injection attacks.
  • Ingestion points: Data enters the agent's context through the search, fetch, and range operations described in sdk-overview.md and quick-start.md.
  • Boundary markers: The documentation does not specify the use of delimiters or 'ignore embedded instructions' warnings when the agent processes retrieved content.
  • Capability inventory: The skill includes write and delete capabilities, such as upsert, delete, and reset (documented in sdk-overview.md), which could be misused if an agent is manipulated by malicious data within a search result.
  • Sanitization: There is no mention of sanitizing or escaping document content before it is parsed or acted upon by the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 11:55 AM
Security Audit — agent-trust-hub — upstash-search-js