CrewAI agent runs code outside a sandbox

HighAV-FW-003Runtime safetyFramework pack

What it finds

A CrewAI agent can execute code, and it is not set to the Docker-sandboxed "safe" mode.

Code the model writes would run directly on the host, with access to its files, network, and secrets.

What an attacker could do: A manipulated agent could run arbitrary code on the machine hosting the crew.

Set code_execution_mode="safe" (Docker sandbox), or turn code execution off for agents that do not need it.

Pro shows the corrected code for your exact file, step by step, and re-scans to confirm the fix. See plans

Finding code: CREWAI_UNSAFE_CODE_EXECUTION
How it detects: Looks for a risky pattern, and stays quiet when a matching safeguard is present.
Confidence: Definite: the risky setting is written plainly in the file.
Scans: Agent source using the named framework (LangChain, CrewAI, AutoGen, OpenAI Agents SDK, Claude Agent SDK, computer-use tools)
OWASP LLM Top 10: LLM08 - Excessive Agency, LLM02 - Insecure Output Handling
NIST AI RMF: MG-2.2
SOC 2: CC6.8, CC7.1

Paste your agent code or MCP config. The free scan runs all 124 checks in about 10 seconds — no account needed.