acad-blocks

Pass

Audited by Gen Agent Trust Hub on Sep 14, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes potentially untrusted AutoCAD drawing data.\n- Ingestion points: AttributeReference.TextString and DynamicBlockReferenceProperty.Value in SKILL.md are used to read text from CAD blocks.\n- Boundary markers: Absent. There are no markers to distinguish between drawing data and agent instructions.\n- Capability inventory: The skill can perform database write operations (attRef.TextString assignment) and write messages to the AutoCAD editor (ed.WriteMessage) in SKILL.md.\n- Sanitization: Absent. The skill does not demonstrate validation of the text read from the drawing.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 14, 2026, 07:41 PM