Data exfiltration path: private data + untrusted input + outbound channel

HighAV-DATA-003Secrets & data exposureCore pack

What it finds

This agent can read private data, takes in untrusted content (user messages, emails, web pages, documents), and can send data out (email, chat, HTTP POST, webhooks) — with no visible destination allowlist or approval step.

Why it matters

This is the combination attackers look for. A single injected instruction in untrusted content can tell the agent to collect private data and send it to the attacker — no code bug required.

What an attacker could do: A crafted email, ticket, or web page could make the agent send customer data, secrets, or files to an outside address.

How to fix it

Break the chain: allowlist where the agent may send data (recipients, domains), require human approval for any outbound message that includes private data, and keep untrusted content away from agents that hold sensitive data.

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

Check details

Finding code: DATA_EXFILTRATION_PATH
How it detects: Looks for a risky pattern, and stays quiet when a matching safeguard is present.
Confidence: Likely: strong signal, but review it in context.
Scans: Any text-based agent source or configuration (JS/TS/Python/JSON/YAML)
OWASP LLM Top 10: LLM01 - Prompt Injection, LLM06 - Sensitive Information Disclosure, LLM08 - Excessive Agency
NIST AI RMF: MS-2.10, MG-2.2
SOC 2: CC6.1, CC6.7, P4.1

Check your agent for this

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

Scan an agent free Run it in CI or your editor