MCP / Tool Description Poisoning

← Agentic AI (OWASP Agentic Top 10)

mcp_tool_poisoning   HARD

Description

A connected tool/MCP server’s description carries hidden instructions that the agent ingests and obeys during planning.

How it works

The agent reads every connected tool/MCP server’s description while planning and trusts it verbatim. A poisoned connector hides a directive inside its otherwise-benign description. With no description integrity check, the planner extracts and obeys the hidden instruction, invoking a tool the user never asked for. A hardened planner would verify description integrity and never treat connector metadata as instructions.

How to exercise it. DVMA is the harness - open this module from the home index and tap the demo action. The screen ships the malicious input and simulates the attacker (e.g. the companion app, crafted intent, or scanned payload) in-process, and the evidence panel prints the proof. The Tools (optional) and Attack inputs below are only needed to reproduce the exploit end-to-end on a real device.

Exploit steps

1. Set up. Build DVMA with a flavor that enables the Agentic AI (OWASP Agentic Top 10) category (e.g. --dart-define-from-file=config/flavors/dev.json) and run on an emulator/simulator you control. The demo needs no external tooling; for the optional on-device reproduction the relevant tools are: promptfoo, Burp Suite, MCP scanner, mitmproxy.
2. Locate the target. From the home index, open MCP / Tool Description Poisoning (mcp_tool_poisoning). The How it works section above describes this module’s specific weakness; the screen states the intended-secure behavior and exposes the vulnerable action.
3. Exploit. Interact with the agent’s memory / tool / sub-agent channel, plant the malicious content, and confirm it influences a later action or is acted on without authentication.
4. Observe the evidence. Trigger the vulnerable action and read the evidence panel - it prints the concrete proof (leaked value, accepted replay, executed payload, or unauthorized result).
5. Contrast with the secure path. Run the module’s secure/hardened action (where provided) and confirm the same attack is rejected - this is what a correct implementation should do.

Tools (optional)

• promptfoo
• Burp Suite
• MCP scanner
• mitmproxy

garak / promptfoo / MCP scanner test the LLM or MCP endpoint behind the app, not the app binary. Point them at the backend model API (find it with mitmproxy / Burp Suite) or a local on-device model server; the in-app demo already exercises the same prompt path.

Attack inputs

Payloads/artifacts you author for the on-device attack. The demo already ships and simulates these in-process (e.g. the malicious companion app / crafted intent is emulated inside the screen), so you only need to craft them to reproduce the exploit on a real device:

• malicious MCP server
• crafted prompt