upwork-invitation-responder

Warn

Audited by Snyk on Jul 23, 2026

Risk Level: MEDIUM
Full Analysis

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

  • Third-party content exposure detected (high risk: 0.85). This skill’s runtime workflow ingests the operating user-pasted “invitation” text (outsider-authored free text from a client) directly as an input, which the agent would then place into LLM context for generating the reply.

Issues (1)

W011
MEDIUM

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

Audit Metadata
Risk Level
MEDIUM
Analyzed
Jul 23, 2026, 09:49 AM
Issues
1
Security Audit — snyk — upwork-invitation-responder