nonica-revit-mcp

Pass

Audited by Gen Agent Trust Hub on Nov 4, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONDATA_EXFILTRATIONEXTERNAL_DOWNLOADSDYNAMIC_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from Revit models, which serves as a potential vector for indirect prompt injection.
  • Ingestion points: Untrusted data such as element parameters, family names, and category keywords are retrieved through tools like get_parameters_from_elementid and get_elements_by_category (documented in SKILL.md and SKILL_LocalLLM.md).
  • Boundary markers: The skill does not define explicit delimiters or instructions to ignore potentially malicious commands embedded within the BIM data.
  • Capability inventory: The agent has the ability to delete model elements (set_delete_elements), modify parameter values (set_parameter_value_for_elements), and export model data to files (create_pdf_export_print), providing a significant impact surface for successful injection.
  • Sanitization: There is no evidence of sanitization or validation of the data retrieved from the Revit model before it is used in subsequent agent reasoning or tool calls.
  • [COMMAND_EXECUTION]: The skill provides tools for making persistent changes to the Revit model environment. This includes set_delete_elements, which permanently removes elements and their dependents, and various set_* tools for modifying element geometry and metadata. The instructions correctly advise the agent to confirm these actions with the user.
  • [DATA_EXFILTRATION]: The create_pdf_export_print and create_cad_export_print tools allow for the extraction of model data to a local folderpath. While intended for standard BIM workflows, this capability could be used to write sensitive model information to unauthorized local or network locations if the agent is manipulated.
  • [EXTERNAL_DOWNLOADS]: The README.md file references a required external component, the "Nonica AI Connector for Revit," hosted at tools.nonica.io. This is a legitimate vendor resource necessary for the skill's primary functionality.
  • [DYNAMIC_EXECUTION]: The skill utilizes a dynamic invocation pattern for element creation via create_tools_invoker. This tool takes a toolName and a dictionary of arguments (argumentIdsAndValues) to perform complex Revit operations, as described in the three-step creation workflow in SKILL.md.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 4, 2026, 05:27 PM