System prompt returned to the client
MediumAV-DATA-006Secrets & data exposureCore packNew in 2.0
What it finds
An API response includes the agent's system prompt or instructions.
Why it matters
System prompts often describe internal tools, rules and sometimes secrets. Returning them hands attackers a map for crafting injections.
What an attacker could do: Anyone calling the API can read the agent's internal instructions.
How to fix it
Never return the system prompt. Return only the answer and the fields the client needs.
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
- SYSTEM_PROMPT_EXPOSED
- How it detects
- Looks for a specific risky pattern in the code or config.
- 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
- LLM06 - Sensitive Information Disclosure, LLM01 - Prompt Injection
- NIST AI RMF
- MS-2.10
- SOC 2
- C1.1
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
- Hardcoded credentials (generic pattern)AV-SECRET-001 · Critical
- Credential or secret placed in the system promptAV-DATA-002 · High
- Data exfiltration path: private data + untrusted input + outbound channelAV-DATA-003 · High
- Credential stored in plain text in the MCP configAV-MCP-102 · High
- All environment variables handed to an MCP serverAV-MCP-111 · High
- Personal data sent to the model without redactionAV-DATA-004 · High