breakscale
Pass
Audited by Gen Agent Trust Hub on Sep 3, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill implements a specialized utility for architectural modeling and load simulation. It utilizes a local Python script,
scripts/breakscale_link.py, to transform user-defined system designs into compressed URL fragments for use with the Breakscale simulator. All operations are local, relying solely on standard Python libraries (zlib, base64, json, argparse) for data processing and file writing. No unauthorized network operations, credential access, or remote code execution patterns were found. - [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted architectural data from chat, design documents, or repositories to generate simulation specifications. While this creates a surface for indirect prompt injection, it is assessed as safe because the processing script performs strict structural validation and numeric type checking on the input JSON data before encoding it into a non-executable format. Ingestion points: Architecture descriptions in chat or repository files. Boundary markers: Absent. Capability inventory: Local file writing and console output for URL generation. Sanitization: Schema validation and type checking within the Python processing script.
Audit Metadata