ClawHub
Back to skill

Security audit

bim-cli

Security checks for vulnerabilities and agentic risk

Overview

The skill is mostly a disclosed BIM CLI helper, but it tells agents to use live web documentation and submit feedback to maintainers without requiring user approval.

Install only if you are comfortable with a third-party BIM CLI operating on your AEC files and host applications. Treat fetched web docs as reference material, and require explicit approval before any feedback submission, Revit write operation, add-in install, MCP registration, or installer execution.

A.I.G

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The skill states the CLI is 'offline' but then instructs the agent to fetch per-driver documentation from a remote website during normal operation. That expands the trust boundary from a local binary to live network content, which can expose prompts, enable content-based prompt injection, or create nondeterministic behavior if the remote docs are changed.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The feedback command transmits information to maintainers and is not necessary for completing the user’s AEC-file task. Even if it 'scrubs' content, it still creates a data exfiltration path and may disclose sensitive workflow details, filenames, errors, or project metadata to a third party.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.