LlamaIndex PandasQueryEngine runs model-written code

HighAV-FW-009Runtime safetyFramework packNew in 2.0

What it finds

The app uses LlamaIndex's PandasQueryEngine, which executes Python the model writes.

Questions from users go to the model, and the model's code runs on your server. Crafted questions have been shown to run arbitrary commands this way.

What an attacker could do: A user question could execute code on the server.

Do not expose PandasQueryEngine to untrusted users. Use fixed, parameterized queries, or run it in an isolated sandbox with no secrets or network.

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

Finding code: LLAMAINDEX_PANDAS_QUERY_ENGINE
How it detects: Looks for a specific risky pattern in the code or config.
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: LLM02 - Insecure Output Handling, LLM08 - Excessive Agency
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.