write-skill

Fail

Audited by Snyk on Nov 16, 2026

Risk Level: CRITICAL
Full Analysis

CRITICAL E004: Prompt injection detected in skill instructions.

  • Potential prompt injection detected (medium risk: 0.30). The skill is a legitimate gstack authoring guide, but its telemetry step drains an artifacts-sync queue to an unspecified destination and says it "ALWAYS RUN" even in plan mode, which is undisclosed outbound data movement rather than a malicious action.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

  • Potentially malicious external URL detected (medium risk: 0.30). The decisive boundary is that diet.md tells the agent to run scripts/eval.mjs from the upstream nicobailon/skill-diet repository at commit 5f41df1 when the shared eval is unavailable, so executed remote code is an operational dependency; the publisher is an individual's repository, which is unofficial, and the bare commit reference is not an enforced digest check, so the tier is medium.

Issues (2)

E004
CRITICAL

Prompt injection detected in skill instructions.

W012
MEDIUM

Unverifiable external dependency detected (runtime URL that controls agent).

Audit Metadata
Risk Level
CRITICAL
Analyzed
Nov 16, 2026, 08:17 PM
Issues
2