redis-development

Warn

Audited by Snyk on May 12, 2026

Risk Level: MEDIUM
Full Analysis

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

  • Third-party content exposure detected (high risk: 0.80). The skill's AGENTS.md and SKILL.md explicitly show using LangCache (Redis Cloud) via its REST/SDK (e.g., POST /v1/caches/{cacheId}/entries/search and lang_cache.search) and instruct agents to read and use cached responses (result[0]["response"]) directly, which are third-party-managed/user-generated entries that could embed instructions and thus influence the agent's actions.

Issues (1)

W011
MEDIUM

Third-party content exposure detected (indirect prompt injection risk).

Audit Metadata
Risk Level
MEDIUM
Analyzed
May 12, 2026, 10:26 AM
Issues
1