solidcraft-editing
Pass
Audited by Gen Agent Trust Hub on Oct 31, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted design files (.sc, .solidcraft) through the 'open' tool, which could potentially contain malicious instructions embedded within the design metadata or parameters.
- Ingestion points: Design files are loaded into the agent's context using the 'open' command as described in SKILL.md.
- Boundary markers: The instructions lack specific boundary markers or warnings to disregard potential instructions found within the design data.
- Capability inventory: The skill is equipped with file system operations (open, save) and structural modification tools (parameters.change, timeline.edit) provided by the MCP server.
- Sanitization: There is no evidence of content sanitization or instruction filtering for the data read from the design files.
Audit Metadata