harmony-progression
Pass
Audited by Gen Agent Trust Hub on Nov 13, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is a collection of reference documents and guidelines. It contains no instructions that attempt to bypass AI safety filters or exfiltrate sensitive information. All external references point to official vendor-owned resources.
- [INDIRECT_PROMPT_INJECTION]: The skill outlines how an agent should assist in processing external game data (saves). It explicitly requires the implementation of security measures such as strict field parsing, value clamping, and output sanitization to prevent attacks from malicious save files.
- Ingestion points: Data retrieved from browser storage (
localStorage,IndexedDB) and uploaded JSON files. - Boundary markers: Direct rules to "Never trust a save" and to parse fields individually rather than merging objects.
- Capability inventory: The skill is scoped to web game logic and does not involve risky shell commands or unauthorized network access.
- Sanitization: Instructions include mandatory character filtering, length capping, and the use of safe DOM APIs (textContent) to prevent XSS.
Audit Metadata