game-development
Pass
Audited by Gen Agent Trust Hub on Oct 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill implements various systems (Quest, Dialogue, Localization, Multiplayer, Save/Load) that ingest external data from JSON files, local storage, and network streams. These represent potential attack surfaces where malicious data could attempt to influence the agent's behavior.
- Ingestion points: Data is loaded from local JSON files in
leveleditor.mdandinventory.md, retrieved from Cloud Firestore insaveload.md, and received via WebSocket messages inmultiplayer.md. - Boundary markers: No specific boundary markers or 'ignore embedded instructions' warnings are present in the code snippets for processing external strings.
- Capability inventory: The skill patterns include local file writing (
leveleditor.md,saveload.md) and network communications via WebSockets and Firebase (saveload.md,multiplayer.md). These are standard functionalities for the intended game development purpose. - Sanitization: The provided patterns rely on standard
jsonDecodeand basic string interpolation with placeholders (e.g., inlocalization.md), which does not provide protection against prompt injection if the source data is compromised.
Audit Metadata