skills/akillness/jeo-skills/agent-benchmarking/Snyk

agent-benchmarking

Warn

Audited by Snyk on Oct 9, 2026

Risk Level: MEDIUM

Full Analysis

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

Third-party content exposure detected (medium risk: 0.30). The workflow references SWE-bench which involves reading GitHub issues, an outsider-authored source.

Issues (1)

W011

MEDIUM

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

Audit Metadata

Risk Level

MEDIUM

Analyzed

Oct 9, 2026, 09:58 AM

Issues

1