codex-subagent
Warn
Audited by Snyk on Aug 10, 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 runtime workflow delegates sub-agent work by taking the user-composed prompt (free text) and piping it directly into
codex exec --full-autovia the Shell tool (echo "" | codex exec ... "<prompt>"), meaning outsider-authored text can reach the LLM without selecting a specific trusted item first.
Issues (1)
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
Audit Metadata