Web pages loaded into the knowledge base without checks
MediumAV-DATA-008Runtime safetyCore packNew in 2.0
What it finds
Web pages are fetched and stored in the vector database with no source allowlist or cleaning step.
Why it matters
Anything on those pages — including hidden instructions — becomes "trusted" knowledge the agent retrieves later. This is RAG poisoning.
What an attacker could do: A web page author can plant instructions that fire whenever the agent retrieves that page.
How to fix it
Only ingest from an allowlist of trusted domains, strip hidden text and markup, and treat retrieved text as data in the prompt.
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
- RAG_UNTRUSTED_WEB_INGEST
- 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, LLM03 - Training Data Poisoning
- NIST AI RMF
- MS-2.3, MP-4.1
- SOC 2
- CC7.1
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
- Command injection riskAV-RUNTIME-004 · Critical
- Unsafe dynamic code executionAV-RUNTIME-005 · Critical
- Privileged or host-mounted container configurationAV-RUNTIME-007 · Critical
- SQL query built from unescaped inputAV-RUNTIME-009 · Critical
- Unsafe deserialization of untrusted dataAV-RUNTIME-010 · Critical
- Prompt injection surface detectedAV-RUNTIME-002 · High