Agent sends emails or messages without review
MediumAV-HUMAN-005Human oversightCore packNew in 2.0
What it finds
A tool lets the agent send emails, Slack messages or texts directly, with no draft or approval step.
Why it matters
Messages go out in your name and cannot be unsent. Injected instructions commonly use "send this to…" to leak data or phish contacts.
What an attacker could do: The agent can contact anyone in your name and include any data it can read.
How to fix it
Have the agent create drafts, and require a person to approve before sending — especially to new recipients.
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
- OUTBOUND_MESSAGE_NO_REVIEW
- 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
- LLM08 - Excessive Agency, LLM06 - Sensitive Information Disclosure
- NIST AI RMF
- MG-2.2
- SOC 2
- CC6.1
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.
Related checks
- Agent can move money with no amount limitAV-HUMAN-003 · Critical
- Missing human-in-the-loop gatesAV-HUMAN-001 · High
- Built-in safety controls turned offAV-HUMAN-002 · High
- Destructive tool with no confirmationAV-HUMAN-004 · High
- AI editor set to run every tool without askingAV-FW-013 · High
- OpenAI Agents SDK agent with tools but no guardrailsAV-FW-006 · Medium