interview-prep

Warn

Audited by Snyk on Apr 19, 2026

Risk Level: MEDIUM
Full Analysis

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

  • Third-party content exposure detected (high risk: 1.00). The SKILL.md explicitly instructs running TinyFish agents to scrape live public, user-generated content from Glassdoor, Blind, and Reddit (see "Step 2 — Parallel scraping" agent commands), so the agent will ingest and act on untrusted third‑party reports that can influence its outputs.

Issues (1)

W011
MEDIUM

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

Audit Metadata
Risk Level
MEDIUM
Analyzed
Apr 19, 2026, 01:45 PM
Issues
1