antigravity-skill-orchestrator
Pass
Audited by Gen Agent Trust Hub on Oct 30, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: Fetches a master skill catalog from a remote GitHub repository to discover available tools.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from a remote catalog and external memory storage, creating an ingestion surface.
- Ingestion points: Remote catalog at
https://raw.githubusercontent.com/sickn33/agentic-awesome-skills/main/CATALOG.mdand outputs from theagent-memory-mcptool. - Boundary markers: Absent; the instructions do not specify delimiters to isolate external data from the agent's primary instructions.
- Capability inventory: The orchestrator itself does not execute shell commands or sensitive file operations, but it manages the selection and invocation of other skills which may have those capabilities.
- Sanitization: Absent; no explicit validation or escaping of retrieved content is performed before interpolation into the prompt context.
Audit Metadata