antigravity-workflows
Pass
Audited by Gen Agent Trust Hub on Feb 19, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [Indirect Prompt Injection] (LOW): Architectural surface identified. 1. Ingestion points: docs/WORKFLOWS.md and data/workflows.json (SKILL.md). 2. Boundary markers: Absent. 3. Capability inventory: Invokes secondary skills such as go-playwright and security-audit-web-app (SKILL.md). 4. Sanitization: Absent. The skill executes actions based on external data which lacks boundary markers.
- [Metadata Poisoning] (SAFE): Frontmatter claim 'risk: none' in SKILL.md was disregarded as an authoritative safety statement.
Audit Metadata