Prompt Injection Welcome Mat
What it is
LLM input, retrieval, or tool paths are present without visible prompt-injection tests or policy boundaries.
This usually appears when model input, retrieval, or tool calls are added faster than capability boundaries and adversarial tests.
How to fix it
Add adversarial prompt tests, tool-call policy checks, source isolation, and refusal paths for untrusted instructions.
- 1Add adversarial prompt tests for retrieval and tool paths.
- 2Check tool-call policy before executing actions.
- 3Isolate untrusted source text and define refusal paths.
How VibeAudit checks for this
VibeAudit checks this with SAST_LOCAL_STRUCTURAL AI/tool route evidence and AI_CODING_AGENT_STATIC indicators for agent or IDE autorun files when those public or local files are in scope. The free tier may show the surface name only, without private evidence or remediation details. Metadata-only surface maps do not create missing-control findings. Unsupported private repositories, authenticated behavior, destructive workflows, account configuration, and business logic remain unverified unless the scan includes matching local, imported, or authorized runtime evidence.
Related failure modes
Check your public launch surface and keep unsupported scope visible.
Run Quick Scan