JARVIS runs entirely on your local machine — it has no servers, no cloud endpoints, and no telemetry. Security concerns mainly relate to the local attack surface: the FastAPI backend port, tool permission tiers, and credential handling.

Please do not open a public GitHub issue for security vulnerabilities.

Use GitHub's private vulnerability reporting to report a vulnerability confidentially. You can expect an acknowledgement within 72 hours and a fix or mitigation plan within 14 days.

Include:

• A clear description of the vulnerability
• Steps to reproduce or a proof-of-concept
• Potential impact (what an attacker could do)
• Your suggested fix, if you have one

• Backend port (:7341) — the FastAPI server listens on 127.0.0.1 by default. Do not expose it to the internet or an untrusted network.
• Tool permission tiers — SAFE / CAUTION / DANGEROUS tiers gate destructive operations. Review config.yaml → permission_mode before enabling auto or bypass on a sensitive machine.
• Credentials — API keys live in .env (git-ignored). Never commit .env. The backend reads keys only at startup; they are never logged or transmitted.
• Generated media — files in data/generated/ are served locally at /api/files/generated/.... The path-traversal guard in routes.py restricts access to the allowlisted roots — do not modify this without review.

• Vulnerabilities in third-party model weights (FLUX, TRELLIS, Whisper, Piper)
• Issues that require physical access to the machine running JARVIS
• Social-engineering attacks on the user