feat(live): deploy bounded simulator supervisor - #245
Merged
Merged
Conversation
4 tasks
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The Phase 3 API can queue simulator requests, but production needs a trusted worker with Docker isolation to process them. This adds a dedicated supervisor VM and digest-pinned worker/verifier images to the existing gated production workflow. The API enables factory mode only after the worker proves a valid simulator passes and a network-using candidate fails in the actual sandbox.
The host reuses the existing managed identity, denies inbound traffic, retrieves secrets at startup, and drains bounded work on shutdown. Disabling the factory also disables the service across reboots. CI builds and scans the new image.
Validation: full local Python suite (3,182 passed, 18 skipped), browser suite (23 passed, including sandbox readiness), Ruff/format; independent Python/style/test/security reviews; image build and fixable HIGH/CRITICAL Trivy gate; Azure Bicep compile and successful what-if. The counts include the companion evaluation fixes being delivered separately.
Refs #237 and #219. This delivers rollout infrastructure; #237 remains open until the gated production release, fallback verification, and Pouyan's learner acceptance. No production deployment has run yet.