antigravity-workflows

Pass

Audited by Gen Agent Trust Hub on Mar 11, 2026

Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
  • [PROMPT_INJECTION]: The skill architecture creates a surface for indirect prompt injection by consuming external data to drive orchestration logic. 1. Ingestion points: The skill reads instructions and metadata from docs/WORKFLOWS.md and data/workflows.json. 2. Boundary markers: No delimiters or instructions to ignore embedded commands are specified in the orchestration logic. 3. Capability inventory: The skill triggers powerful sub-skills for security audits and browser automation which involve network and system access. 4. Sanitization: There is no validation or filtering of the external workflow content before interpretation.
  • [NO_CODE]: The skill consists of documentation and orchestration instructions in Markdown format and does not include any scripts or binary files.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 11, 2026, 07:04 PM