skills/danicat/skills/ebitengineer/Gen Agent Trust Hub

ebitengineer

Pass

Audited by Gen Agent Trust Hub on Oct 14, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill consists of technical documentation and architectural patterns for Ebitengine-based game development. Implementation examples for file persistence in references/wasm_audio_and_persistence.md and references/input_action_mapping.md correctly utilize atomic writes and restricted file permissions (0o600). The architectural patterns promote modularity and fail-fast validation of game assets, which are standard development practices.
  • [INDIRECT_PROMPT_INJECTION]: The skill provides code templates that ingest and process external data, establishing an indirect injection surface. * Ingestion points: Data ingestion occurs in references/input_action_mapping.md (loading JSON input bindings), references/wasm_audio_and_persistence.md (loading JSON game saves), and references/server_architecture.md (accepting JSON score entries via a REST API). * Boundary markers: Explicit boundary markers are not present in the code snippets provided. * Capability inventory: The code templates demonstrate file system interactions via the os package (read, write, rename, and directory creation) and network service management via net/http. * Sanitization: The provided code includes basic validation logic, such as checking for empty strings, positive integer values, and schema version consistency in JSON payloads.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 14, 2026, 02:34 AM