feat(integrations): Add VAREK Guardrails + W&B pipeline verification …

[kwdoug63]

What this example demonstrates

This example shows how to verify that AI/ML pipeline code is being executed under deterministic kernel-level isolation (seccomp-bpf + cgroups + wallclock), with all results logged to W&B for audit and reproducibility.

VAREK Guardrails is an open-source runtime sandbox that wraps subprocess.Popen with a seccomp-bpf filter, a cgroups v2 subtree (memory/CPU/PIDs), and a wallclock killer. This integration demonstrates three patterns that map directly to W&B-supported workflows.

What's included

examples/varek-guardrails/
├── README.md
├── requirements.txt
├── verification_artifact.py # 5-test verification battery → W&B Artifact
├── telemetry_stream.py # subscribe_telemetry → W&B run logs
├── benchmark_suite.py # 8-payload regression suite → W&B Table
└── scripts/
└── with-cgroup.sh # idempotent cgroup subtree wrapper

Why W&B users benefit

For teams running untrusted-code execution (LLM agents, eval harnesses, code-as-data pipelines), this provides a verifiable audit trail: every claim about isolation is backed by a logged run, and the runs are reproducible by any reviewer with the same versions installed. The Artifact pattern lets downstream stages depend on the verification result; the Table pattern catches policy regressions automatically.

Tested against

• varek-guardrails 1.1.1
• wandb 0.26.1
• weave 0.50+
• Ubuntu 24.04 (Linux 6.x, cgroups v2)

Live verification runs

End-to-end runs from the development environment (sober-agents entity):

Script	Run	Result
verification_artifact.py	fast-jazz-2	5/5 PASS, artifact logged
telemetry_stream.py	skilled-bird-2	3 payloads, ~2475 audit events captured
benchmark_suite.py	fluent-butterfly-2	8/8 OK, match_rate=100%

Notes for reviewers

• Requires Linux with cgroups v2 (kernel 4.5+, all major distros from 2019 onward)
• The wrapper script (scripts/with-cgroup.sh) is required because systemd's Delegate= directive does not apply to .slice units, so cgroup subtree controls must be set explicitly before invocation
• The benchmark suite includes a known_allowance category for os.system passthrough — preserved deliberately as a regression test against future policy tightening in varek_guardrails
• The telemetry callback uses a PID guard + queue.Queue buffer to remain safe in post-fork-pre-exec contexts (audit hooks fire in forked children; calling wandb.log directly there breaks the seccomp arming)
• Directory placement (examples/varek-guardrails/) is a suggestion — happy to move under a different parent (e.g. examples/security/ or examples/integrations/) based on maintainer preference

Happy to iterate on naming, scope, or any of the patterns.

[socket-security]

Review the following changes in direct dependencies. Learn more about Socket for GitHub.

Diff	Package	Supply Chain Security	Vulnerability	Quality	Maintenance	License
	weave@​0.52.37

View full report