scene-architecture
Pass
Audited by Gen Agent Trust Hub on Aug 8, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill consists entirely of instructional markdown text. It contains no executable scripts (Python, JavaScript, or shell) and no binary files.
- [SAFE]: No obfuscation techniques, such as Base64, zero-width characters, or homoglyphs, were detected in the content.
- [SAFE]: The referenced URL (https://writecalliope.ink/docs/mcp) is a plain-text link to the author's documentation for an MCP connector. This is a legitimate vendor resource used to describe how the skill can optionally interact with user-selected manuscript files in a read-only capacity.
- [SAFE]: While the skill defines an attack surface for indirect prompt injection by processing user writing samples and manuscript files, it lacks the dangerous capabilities (such as file-write operations, network exfiltration, or command execution) required to weaponize such an injection. The instructions explicitly state the skill is read-only and never writes or rewrites prose.
Audit Metadata