rollback-workflow-builder
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The GitHub Action workflow templates provided in the skill ingest untrusted input via workflow_dispatch events. Specifically, the 'reason' input is interpolated directly into the markdown body of a GitHub Issue without sanitization or boundary markers. This creates a surface for indirect prompt injection where an attacker could provide a malicious string to influence the behavior of agents or platforms processing the resulting issue.\n
- Ingestion points: SKILL.md (Manual Rollback Workflow section)\n
- Boundary markers: Absent (input is placed directly into a markdown template)\n
- Capability inventory: GitHub API interaction (issue creation) and shell script execution\n
- Sanitization: Absent
Audit Metadata