Files
agentbox/SECURITY.md
T
Yash Shah a21089184b feat(v0.4): enforce per-run memory_mb via RLIMIT_AS
Context:
- Timeout alone does not stop runaway memory allocations in agent-generated code.
- Docs understated that subprocess is not production isolation.

Test Plan:
- docker compose run --rm test
- tests/test_api.py::test_memory_limit_kills_huge_allocation

Additional Notes:
- limits.memory_mb uses resource.RLIMIT_AS via preexec_fn (Linux Docker).
- Adoption: badges, SECURITY.md, CONTRIBUTING, author/urls.
2026-09-14 14:00:42 +05:30

878 B

Security Policy

Reporting a vulnerability

Email yash376351@gmail.com with the repo name, a short description, and steps to reproduce. Please do not open a public issue for exploitable findings until we have had a reasonable chance to respond.

Threat model (honest)

agentbox executes untrusted code on your machine via a subprocess sandbox.

  • The current backend is subprocess + optional RLIMIT_AS, not gVisor/Firecracker. Do not expose it to the public internet as-is.
  • Credential stripping and timeouts reduce accidents; they are not a kernel isolation story.
  • limits.memory_mb caps address space via RLIMIT_AS — it is not a cgroup memory controller and behavior differs by OS.
  • Snapshots persist workspace files on disk under AGENTBOX_SNAPSHOT_DIR.
  • Prefer running the API only on trusted networks, behind auth you add yourself.