api-database-vercel-kv

Pass

Audited by Gen Agent Trust Hub on Sep 20, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill enables the agent to pull data from an external Redis database into its context, creating a surface for indirect prompt injection.
  • Ingestion points: Data is retrieved using methods like redis.get, redis.hgetall, and redis.mget (found in SKILL.md, examples/core.md, and reference.md).
  • Boundary markers: Absent; the provided code patterns do not include delimiters or instructions for the agent to ignore potentially malicious commands embedded in the database content.
  • Capability inventory: The skill provides the agent with full read, write, and deletion capabilities for the Redis instance (found in SKILL.md and examples/core.md).
  • Sanitization: Absent; the documentation does not demonstrate sanitization or schema validation for data retrieved from the remote store.
  • [EXTERNAL_DOWNLOADS]: The skill utilizes official and well-known packages, including @upstash/redis, @vercel/kv, and @upstash/ratelimit, to facilitate database connectivity and management.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 20, 2026, 03:57 PM
Security Audit — agent-trust-hub — api-database-vercel-kv