ci: add GitHub Actions CI workflow with lint, test, and image build by drew · Pull Request #1 · NVIDIA/OpenShell

drew · 2026-02-27T05:24:39Z

Summary

Adds .github/workflows/ci.yml with lint, rust-test, python-test, and CI image build jobs
Adds build/scripts/ci-image.sh for local + CI docker image builds
Adds docker:buildx:setup mise task for multi-arch remote BuildKit configuration
Updates docker:build:ci mise task to use the new script

…tall steps

…action Use docker/setup-buildx-action@v3 wrapped in a local composite action to create the multi-arch remote BuildKit builder. This ensures automatic cleanup of builder nodes when the job finishes, preventing leaked BuildKit instances on self-hosted runners.

Declare the ARG before mise install so authenticated GitHub API requests avoid rate limits when fetching tool versions.

Implement the final integration for GPU passthrough support: - Add put_vfio_device() API function for PUT /vfio/{device_id} - Wire prepare_gpu_attachment() into do_create_inner() - Enable supports_gpu in Firecracker capabilities - Update spec compatibility checks and tests This completes blockers NVIDIA#1 (GPU attachment not wired) and NVIDIA#2 (VFIO device attachment API not implemented). Signed-off-by: OpenCode Agent <opencode@nvidia.com>

feat(podman): add Podman support as alternative container runtime

docs(podman): Add macOS installation guide and Podman build compatibility

1. Separate client_id from audience (Critical/High, findings NVIDIA#1/NVIDIA#3): - Add oidc_audience field to GatewayMetadata separate from oidc_client_id - Bootstrap stores client_id and audience independently - Fixes the conflation that made the docs overstate provider portability 2. Skip list risk documentation (Critical, finding NVIDIA#2): - The skip list is by design for sandbox supervisor RPCs which use SSH handshake secrets. Noted for future hardening with per-sandbox credentials. 3. Reject partial-empty RBAC config (Medium, finding NVIDIA#4): - AuthzPolicy::validate() rejects configs where only one of admin_role/user_role is set - Server validates at startup before accepting requests - Prevents silently opening admin endpoints to any authenticated user 4. Preserve refresh token on refresh (Medium, finding NVIDIA#5): - oidc_refresh_token() keeps the old refresh_token when the server doesn't return a new one, per OAuth 2.0 spec 5. Additional concerns: - Percent-decode callback query parameters (code, state, error) - Drop scope=openid from client_credentials flow - Use /dev/urandom for PKCE verifier/state on Unix - Validate discovery issuer matches configured issuer (both server and CLI) to prevent SSRF/misdirection - Wire RBAC config (rolesClaim, adminRole, userRole) through Helm values and statefulset template

1. Critical: sandbox RPCs now require x-sandbox-secret header (NVIDIA#1): - Split skip list into unauthenticated (health/reflection) and sandbox-secret (supervisor RPCs) categories - Sandbox RPCs require x-sandbox-secret header matching the server's SSH handshake secret - Sandbox gRPC client injects the secret via SandboxSecretInterceptor - Unauthenticated network callers can no longer reach sandbox RPCs 2. High: complete client_id vs audience separation (NVIDIA#2): - Add --oidc-audience flag to gateway add (separate from --oidc-client-id) - Store oidc_audience in gateway metadata independently - Thread audience through bootstrap metadata correctly 3. Medium: propagate RBAC settings through bootstrap (NVIDIA#3): - Add oidc_roles_claim, oidc_admin_role, oidc_user_role to DeployOptions - Pass OIDC_ROLES_CLAIM, OIDC_ADMIN_ROLE, OIDC_USER_ROLE env vars through Docker container → cluster-entrypoint.sh → HelmChart manifest - Full provider-neutral RBAC config now works via gateway start 4. Low: cross-platform CSPRNG (NVIDIA#4): - Replace /dev/urandom + RandomState fallback with getrandom crate - Works correctly on all platforms (Linux, macOS, Windows)

1. High: UpdateConfig dual-auth for sandbox policy sync (NVIDIA#1): - Add DUAL_AUTH_METHODS list for RPCs that accept either Bearer token (CLI users) or sandbox secret (supervisor) - UpdateConfig now works for both CLI policy mutations and sandbox startup policy sync in OIDC deployments 2. Medium: end-to-end client_id vs audience separation (NVIDIA#2): - Thread optional audience parameter through oidc_browser_auth_flow and oidc_client_credentials_flow - gateway login reads oidc_audience from metadata and passes it to the auth/token requests - Adds audience parameter to authorization URL for providers that require it (Entra ID) 3. Medium: gateway start exposes full OIDC RBAC config (NVIDIA#3): - Add --oidc-client-id, --oidc-roles-claim, --oidc-admin-role, --oidc-user-role flags to gateway start - Thread all values through gateway_admin_deploy into DeployOptions - Stop hardcoding "openshell-cli" as client_id in bootstrap metadata

…drafts Three things in one commit, all surfaced by running the demo end-to-end against a real gateway and finding the agent had to draft a broader second proposal. 1. /wait race fix. Previously /wait returned `approved` the moment it observed the gateway's chunk status flip, but the local supervisor reloads policy on its own poll cycle (~10s in practice). The agent's retry would race the reload and hit the still-old policy, getting denied. Codex then drafted a broader rule and re-submitted — sound agent behavior, but not what /wait should provoke. Now /wait captures the local policy version at start, and after observed-approved waits for the supervisor to load a strictly-newer version before returning. Bounded by the caller's deadline; best-effort return if the deadline elapses without the version bumping. Two new unit tests pin the happy path and the deadline-clamped fallback. 2. demo.sh auto-approve loop. Replaces approve_when_pending + wait_for_agent with one approve_pending_until_agent_exits function that keeps watching for pending chunks and approving them until the agent process exits (or the configured timeout). Defense in depth against future redraft scenarios for any reason; today (post-fix#1) the agent should only submit one proposal per task, but we don't want to hang silently if it does submit more. 3. UX. Step headers now carry "[t+1.2s]" relative timestamps so reading the run output makes latency visible (the demo's whole point is the wait is cheap — surface that). A spin_wait helper renders an ASCII spinner during the watch loop so the demo never looks frozen on a TTY. Falls back to plain sleep on non-TTY contexts. Closes the race condition diagnosed from the trace timing where the gateway approved at t+0, sandbox observed at t+0.3s, but the supervisor didn't load v2 until t+9.4s — well after the agent had already retried and been denied. Signed-off-by: Alexander Watson <zredlined@gmail.com>

johntmyers · 2026-05-12T15:55:57Z

smoke test - ignore

…drafts Three things in one commit, all surfaced by running the demo end-to-end against a real gateway and finding the agent had to draft a broader second proposal. 1. /wait race fix. Previously /wait returned `approved` the moment it observed the gateway's chunk status flip, but the local supervisor reloads policy on its own poll cycle (~10s in practice). The agent's retry would race the reload and hit the still-old policy, getting denied. Codex then drafted a broader rule and re-submitted — sound agent behavior, but not what /wait should provoke. Now /wait captures the local policy version at start, and after observed-approved waits for the supervisor to load a strictly-newer version before returning. Bounded by the caller's deadline; best-effort return if the deadline elapses without the version bumping. Two new unit tests pin the happy path and the deadline-clamped fallback. 2. demo.sh auto-approve loop. Replaces approve_when_pending + wait_for_agent with one approve_pending_until_agent_exits function that keeps watching for pending chunks and approving them until the agent process exits (or the configured timeout). Defense in depth against future redraft scenarios for any reason; today (post-fix#1) the agent should only submit one proposal per task, but we don't want to hang silently if it does submit more. 3. UX. Step headers now carry "[t+1.2s]" relative timestamps so reading the run output makes latency visible (the demo's whole point is the wait is cheap — surface that). A spin_wait helper renders an ASCII spinner during the watch loop so the demo never looks frozen on a TTY. Falls back to plain sleep on non-TTY contexts. Closes the race condition diagnosed from the trace timing where the gateway approved at t+0, sandbox observed at t+0.3s, but the supervisor didn't load v2 until t+9.4s — well after the agent had already retried and been denied. Signed-off-by: Alexander Watson <zredlined@gmail.com>

* feat(policy): plumb chunk_ids and rejection_reason through proposal pipeline Prereq plumbing for the agent revise-and-resubmit loop. Two narrow additive proto changes unblock the upcoming /wait endpoint (#1092), prover validation badge (#1097), and reject --guidance surfaces (#1098). - SubmitPolicyAnalysisResponse: add accepted_chunk_ids so the in-sandbox agent gets handles to watch its proposals. Surfaced through the typed grpc_client wrapper and policy.local's POST /v1/proposals 202 body. Closes#1094. - PolicyChunk + StoredDraftChunk + DraftChunkPayload: add validation_result (gateway prover verdict, populated by #1097) and rejection_reason (operator free-form text). Both plain strings; no enums, no parsing on the read path. Closes#1096. - RejectDraftChunk now persists the existing reason field into the chunk's rejection_reason so it round-trips back to the agent via GetDraftPolicy. UndoDraftChunk clears it on the way back to pending so consumers cannot read a stale guidance string from a prior reject -> re-approve -> undo cycle. Whole surface stays gated behind agent_policy_proposals_enabled. Two focused tests cover the round-trip and the undo-clears guarantee. Signed-off-by: Alexander Watson <zredlined@gmail.com> * feat(sandbox): add /v1/proposals/{id} and /wait long-poll to policy.local The agent feedback channel back from policy.local. Two new routes let the in-sandbox agent learn its proposal's outcome on a single blocking HTTP call — zero LLM tokens during the wait. - GET /v1/proposals/{chunk_id} returns the chunk's current state in one gateway call. - GET /v1/proposals/{chunk_id}/wait?timeout=<s> blocks until the chunk transitions out of pending. Default 60s, clamped [1, 300]. Agent re-issues on timeout to extend. Response carries the chunk's status plus the two feedback fields shipped in the prereq commit: rejection_reason (free-form reviewer text) and validation_result (gateway prover verdict, empty until #1097). On timeout: same shape with timed_out: true so the agent can disambiguate without parsing. Wait handler short-polls GetDraftPolicy every 1s inside the request with a tokio::time::Instant deadline. One gateway connection is opened per request and reused across all polls, so a 60s wait does one TLS handshake instead of sixty. A future commit can swap the loop body for a tokio::sync::broadcast driven by a watcher task — the agent-visible contract (URL, query, response shape) is independent of the polling implementation. All routes stay behind agent_policy_proposals_enabled. Closes#1092. Signed-off-by: Alexander Watson <zredlined@gmail.com> * docs(sandbox): teach policy_advisor skill the wait + redraft loop The agent-facing instructions for the feedback loop. The endpoints exist; this is the doc that makes them usable. policy_advisor.md gains: - API entries for GET /v1/proposals/{chunk_id} and /wait?timeout=<s>, including the field semantics (status, rejection_reason, validation_result, timed_out). - A note on the submit response's accepted_chunk_ids / rejection_reasons split so the agent handles partial acceptance. - Step 6 saves the chunk_ids and addresses any submit-time rejections before waiting. - Step 7 walks the four wait outcomes: approved (retry, with the honest "may still fail" caveat), rejected (read rejection_reason AND validation_result; address whichever has content), still-pending with timed_out (re-call), non-2xx (surface, do not retry). skills.rs gains two assertions on the skill content so a future edit cannot drop the wait endpoint or the rejection_reason directive silently. Closes#1095. Signed-off-by: Alexander Watson <zredlined@gmail.com> * test(policy-advisor): add end-to-end smoke for the agent feedback loop A focused smoke that exercises the new policy.local /wait endpoint on a live gateway + sandbox, separate from the existing no-LLM regression harness (which still drives the OLD retry-with-bash-loop recovery pattern). Two flows: - Flow A — approve-and-retry: agent submits, /wait blocks, host runs `openshell rule approve`, /wait returns status=approved. Confirms the happy path round-trip latency. - Flow B — reject-with-guidance: agent submits, /wait blocks, host runs `openshell rule reject --reason "..."`, /wait returns status=rejected with the exact reviewer text in rejection_reason. Confirms the free-form guidance contract round-trips through the agent feedback channel. No GitHub credentials needed — proposals are synthetic and never trigger outbound traffic. Both flows expect agent_policy_proposals_enabled=true and a running gateway. Adds three cases to sandbox-runner.sh: submit-test-proposal (no GH deps), proposal-status, proposal-wait. The existing put-file and submit-proposal cases used by test.sh are untouched. Signed-off-by: Alexander Watson <zredlined@gmail.com> * fix(policy-advisor): surface real CLI errors from wait-smoke preflight The preflight piped openshell's stderr to /dev/null and relied on jq to default the missing setting key to "<unset>", but under `set -euo pipefail` a non-zero exit from openshell makes the whole pipeline fail and the command substitution exits the script silently before the intended fail() message can print. Capture stderr explicitly, check the CLI exit code, and surface the real error plus the expected fix (port-forward + gateway add + select) when the CLI cannot reach the gateway. Signed-off-by: Alexander Watson <zredlined@gmail.com> * fix(policy-advisor): pass --json to settings get in wait-smoke preflight `openshell settings get --global` defaults to a human-readable table; jq cannot parse it and the preflight died with a numeric-literal error. Pass --json so jq gets actual JSON. Also touched up the suggested recovery commands in the preflight error to match the real CLI shape (`gateway add <endpoint> --name <name>` and the env-var override warning). Signed-off-by: Alexander Watson <zredlined@gmail.com> * fix(policy): dedup draft chunks only in mechanistic mode; return effective id The smoke harness for the agent feedback loop caught a real bug in the gateway: SubmitPolicyAnalysis's response carried a chunk_id that was never persisted whenever the SQL ON CONFLICT path fired. Two failure modes, both load-bearing: - Agent-authored proposals targeting the same host/port/binary (e.g. the redraft-after-rejection loop) silently folded into one row and any RejectDraftChunk by the new chunk_id failed with "chunk not found." Latent since #1151, surfaced by #1094 returning chunk_ids. - Mechanistic mode had the same class of bug — the dedup fold-in is the intended behavior there, but the response still advertised the newly-generated UUID instead of the existing row's id. Less visible because no current caller reads mechanistic chunk_ids back, but the proto contract was violated either way. Fix in three parts: - put_draft_chunk now takes Option<&str> dedup_key explicitly and returns the effective row id (via RETURNING). None binds NULL to the dedup_key column, which bypasses the partial-index ON CONFLICT path entirely. Caller-decides semantics replace store-side magic. - handle_submit_policy_analysis picks dedup_key per chunk using an allowlist (only "mechanistic" dedups) and pushes the returned effective_id to accepted_chunk_ids. New modes default to no-dedup so a misconfigured caller cannot silently lose proposals. - The two-copy draft_chunk_dedup_key helper consolidated to one observation_dedup_key in policy_store.rs with a doc comment. Tests: - agent_authored_submits_for_same_endpoint_do_not_dedup pins the redraft-loop contract: two intentional submissions with the same host/port/binary get distinct chunk_ids, both findable via GetDraftPolicy, both rejectable by id. - mechanistic_submits_for_same_endpoint_dedup_into_one_chunk locks in the observation-mode dedup AND asserts both submits return the same effective_id — would have caught the deeper bug. Proto: SubmitPolicyAnalysisRequest.analysis_mode doc updated to describe the actual semantics (mechanistic dedups, agent_authored and unknown modes do not). Signed-off-by: Alexander Watson <zredlined@gmail.com> * docs(examples): retarget policy-management demo at the /wait endpoint The narrated demo (examples/agent-driven-policy-management) has been the public face of this feature since #1151. Its agent prompt told Codex to retry the original PUT every few seconds for up to 120 seconds — a polling workaround for the missing /wait endpoint that this branch shipped. Update the demo to exercise /wait so the canonical reading of the feature reflects the actual UX win. - agent-task.md: step 4 is now "call /wait, branch on status" with the three outcomes spelled out (approved → retry once; rejected → read rejection_reason and revise or stop; pending+timed_out → re-issue /wait once, do NOT busy-loop or shorten the timeout). Also makes explicit that the demo submits one rule per proposal so accepted_chunk_ids[0] is the safe single id to wait on. - demo.sh: header docstring rewritten as a six-step loop that mirrors the README. narrate_sandbox_workflow drops its parallel numbering and uses bullets (the runtime narration is the agent's sub-actions, not a separate decomposition of the loop). Approve step header and success message now reference /wait waking the agent, not "policy hot-reload retry." - README.md: top-of-file flow expanded from 5 to 6 steps to include the /wait call and chunk_id capture; "Going further" section now describes both regression scripts and the boundary between them (real-GitHub retry vs. synthetic /wait wire test). Slow-path qualifier corrected from "image pull on first run" to "sandbox cold-start (SSH bring-up plus Codex install)". - wait-smoke.sh header rewritten to make it unambiguous this is a regression, NOT a tutorial, with explicit prereq commands instead of prereq descriptions, and a pointer at demo.sh for the narrated story. No code paths change; this is the readability pass. Signed-off-by: Alexander Watson <zredlined@gmail.com> * fix(examples): pass --yes on demo.sh's global setting writes Global setting updates require explicit confirmation in non-interactive mode; demo.sh's enable_agent_proposals and the cleanup restore path were missing --yes and hard-failed the preflight. Pre-existing issue that surfaced now that more of the demo runs through this path. No other behavior change. Signed-off-by: Alexander Watson <zredlined@gmail.com> * feat(sandbox): emit OCSF audit events for policy proposal lifecycle The demo's policy decision trace previously showed only the proxy enforcement story (HTTP:PUT DENIED, CONFIG:LOADED, HTTP:PUT ALLOWED). It was silent about who proposed what or who decided what — the audit-trail receipts for the agent feedback loop were missing. policy.local now emits sandbox-side OCSF events at the observation moments, into the same stream as the existing CONFIG:LOADED: - CONFIG:PROPOSED on submit_proposal acceptance. Per accepted chunk: the message names the chunk_id, target endpoint, L7 method/path, and binary so the trace correlates against the inbox card via chunk_id. - CONFIG:APPROVED on /wait observation of approved status. - CONFIG:REJECTED on /wait observation of rejected status. Carries the reviewer's free-form rejection_reason in the message AND as an unmapped field, both sanitized (control chars stripped, capped at 200 chars with an ellipsis marker). The agent still reads the raw text via GET /v1/proposals/{id}; sanitization is audit-side only, per AGENTS.md's no-secrets-in-OCSF rule. The submit path defends the audit_summaries / accepted_chunk_ids index pairing against a future gateway change that compresses past rejected chunks (the proto doesn't promise 1:1 ordering with the request). Today client-side validation makes the lengths always match; if they don't, the pairing falls back to a generic per-id event rather than mis-attribute. The wait handler's emit site fires once per terminal-status observation. Multiple concurrent waiters on the same chunk would each emit one event; acceptable for single-waiter-per-chunk demos and the right place to dedup is the SIEM. demo.sh's trace filter now surfaces the four CONFIG: events alongside HTTP:PUT, so the trace at the end of every run tells the full story from deny to allow via propose -> approve. wait-smoke.sh's prereq notes recommend redirecting kubectl port-forward output so its "Handling connection for 8090" lines don't bleed into demo narration. Three new unit tests on the sandbox-side helpers — summary builder happy path, fallback, and the rejection_reason sanitizer. Signed-off-by: Alexander Watson <zredlined@gmail.com> * feat(policy): /wait awaits local policy reload; demo auto-approves redrafts Three things in one commit, all surfaced by running the demo end-to-end against a real gateway and finding the agent had to draft a broader second proposal. 1. /wait race fix. Previously /wait returned `approved` the moment it observed the gateway's chunk status flip, but the local supervisor reloads policy on its own poll cycle (~10s in practice). The agent's retry would race the reload and hit the still-old policy, getting denied. Codex then drafted a broader rule and re-submitted — sound agent behavior, but not what /wait should provoke. Now /wait captures the local policy version at start, and after observed-approved waits for the supervisor to load a strictly-newer version before returning. Bounded by the caller's deadline; best-effort return if the deadline elapses without the version bumping. Two new unit tests pin the happy path and the deadline-clamped fallback. 2. demo.sh auto-approve loop. Replaces approve_when_pending + wait_for_agent with one approve_pending_until_agent_exits function that keeps watching for pending chunks and approving them until the agent process exits (or the configured timeout). Defense in depth against future redraft scenarios for any reason; today (post-fix#1) the agent should only submit one proposal per task, but we don't want to hang silently if it does submit more. 3. UX. Step headers now carry "[t+1.2s]" relative timestamps so reading the run output makes latency visible (the demo's whole point is the wait is cheap — surface that). A spin_wait helper renders an ASCII spinner during the watch loop so the demo never looks frozen on a TTY. Falls back to plain sleep on non-TTY contexts. Closes the race condition diagnosed from the trace timing where the gateway approved at t+0, sandbox observed at t+0.3s, but the supervisor didn't load v2 until t+9.4s — well after the agent had already retried and been denied. Signed-off-by: Alexander Watson <zredlined@gmail.com> * fix(sandbox): /wait detects policy reload by content, not the schema version The previous attempt at the /wait-after-approve race fix compared `SandboxPolicy.version` between /wait start and the policy reload — but that field is the *schema* version (constant 1), not a revision counter. Every comparison was `current(1) > baseline(1) == false`, so the wait blocked until the agent's 300s timeout regardless of whether the supervisor had actually reloaded. The demo SSH connection then timed out around the 240s mark. Diagnosed from a live run's OCSF trace: supervisor pulled v2 at +8.5s after approval (CONFIG:LOADED), but the sandbox-side CONFIG:APPROVED that my /wait emits didn't fire until +304s — exactly at the 300s deadline. Fix: compare the whole policy via prost's derived PartialEq. Any field change (network_policies map being the only one that actually mutates today) flips equality. A clone-per-200ms-tick on a few-KB proto is cheap inside the bounded wait window. Tests rewritten to match the new contract: the supervisor-reload fixture now keeps `version: 1` constant and changes `network_policies` contents, mirroring the exact failure mode from the live run. Signed-off-by: Alexander Watson <zredlined@gmail.com> * fix(examples): redact tokens with python literal-string replace, not sed The sed-based redact_log in demo.sh broke when one of the auth tokens contained a character that conflicted with sed's pattern parser ("unterminated substitute pattern" on the Codex JWT). The whole log tail then blanks on failure, hiding the very failure context we're trying to surface. Switch to a python subprocess that takes the tokens via argv and does literal str.replace. No regex, no delimiter games, no truncation. Signed-off-by: Alexander Watson <zredlined@gmail.com> * fix(sandbox): scope /wait reload check to the approved rule Reviewer (John Myers) flagged two failure modes in the prior whole-policy fingerprint approach used by policy.local /wait: - False sleep: when the supervisor reloads between two /wait calls (the skill tells the agent to re-issue on timed_out), the new call snapshots the already-updated policy as baseline and burns the full timeout waiting for a change that never comes. - False wakeup: any unrelated reload (other agent's approval, settings change) flips the diff, but the chunk's actual rule may not be loaded yet — the agent retries and hits policy_denied for no real signal. Replace the diff with rule-coverage. New public helper openshell_policy::policy_covers_rule reuses endpoints_overlap (so it matches add_rule's merge semantics, including the fold-into-existing-key case) plus an L7 allow check on method/path (so an existing endpoint that doesn't yet contain the proposed method doesn't signal coverage). Add policy_reloaded: true|false to the /wait response on approve, with a 500ms floor on the reload-wait phase so approvals arriving near the deadline still get a fair shot at reloaded=true. Update the policy_advisor skill to branch on it: reloaded=true → retry; reloaded=false → re-issue /wait once with timeout=30, then surface to user. Don't loop tightly. Tests: - 9 new unit tests in openshell-policy pinning coverage semantics (L4-only, L7 method gap, fold-into-existing-key, empty binaries). - 4 new tokio tests in policy_local mirroring John's exact scenarios. - wait-smoke.sh asserts policy_reloaded=true on Flow A. Signed-off-by: Alexander Watson <zredlined@gmail.com> * fix(server): make GetDraftPolicy dual-auth so /wait works under OIDC policy.local calls GetDraftPolicy from inside the sandbox supervisor via the sandbox gRPC client, which authenticates with the shared x-sandbox-secret. GetDraftPolicy was listed only in the Bearer-auth scope table (config:read) and was not in SANDBOX_SECRET_METHODS or DUAL_AUTH_METHODS, so OIDC-enabled gateways rejected those calls and the /wait long-poll surfaced gateway_lookup_failed. Local/no-OIDC setups happened to work because the auth check is short-circuited. Add GetDraftPolicy to DUAL_AUTH_METHODS, matching the existing GetSandboxConfig pattern (called by both CLI reviewer surfaces with Bearer and the sandbox supervisor with x-sandbox-secret). Dual-auth short-circuits the scope check for sandbox-secret callers, so the config:read entry in authz.rs continues to gate Bearer-only flows. Mirror the openshell_get_sandbox_config_is_dual_auth assertion for GetDraftPolicy. Note: ssh_handshake_secret is server-wide, not per-sandbox, so a sandbox-secret caller can today name any sandbox in a SubmitPolicyAnalysis request — and now in a GetDraftPolicy request. The exposure is symmetric with the existing SANDBOX_SECRET_METHODS pattern. Filed as a follow-up: per-sandbox secret binding, tracked separately. Signed-off-by: Alexander Watson <zredlined@gmail.com> * fix(ci): address rebased check failures Signed-off-by: Alexander Watson <zredlined@gmail.com> --------- Signed-off-by: Alexander Watson <zredlined@gmail.com>

- Remove dead boolCount function that would fail golangci-lint (#1) - Emit EventAdded for the first watch event instead of EventModified, matching k8s watch semantics (#7) - Add mutex locking to all mock server methods that access the shared sandboxes map, fixing latent race conditions (#12) - Skip HealthCheck integration test that calls an unimplemented stub (#13) - Scope doc.go examples: mark sections for sub-clients not yet available in this PR with "available in a future release" (#4) - Document Config.Timeout/RetryPolicy/Logger and WatchOptions fields as reserved for future use (#2, #6) Signed-off-by: Roland Huß <rhuss@redhat.com>

…VIDIA#2271) * feat(sdk/go): add Go SDK foundation, types, and sandbox client (A) Add the Go SDK module with the full API contract and a working sandbox client as the first vertical slice. All other resource clients are present as stubs returning Unimplemented errors, to be replaced with real implementations in subsequent PRs. Contents: - Module setup (go.mod, Makefile, mise.toml) - All domain types (types/ package) - Full ClientInterface with all sub-client accessors - Shared infrastructure (errors, auth, gRPC connection, logging) - Sandbox client with converter and tests (fully functional) - Stub clients for remaining resources (exec, file, health, provider, profile, config, refresh, policy, service, ssh, tcp) Part of the Go SDK decomposition plan (NVIDIA#2270). Implements NVIDIA#2044. * fix(sdk/go): address review feedback on PR NVIDIA#2271 - Make scheme parsing drive transport selection: http:// uses plaintext gRPC, https:// or no scheme uses TLS. Add regression tests. - Add Resources and DriverConfig fields to SandboxTemplate and update both converter directions (SandboxFromProto/SandboxSpecToProto). - Regenerate proto bindings from current canonical proto sources to eliminate drift (SigV4/MCP fields, params matchers, reserved fields). - Run gofmt/goimports on all handwritten Go files. Signed-off-by: Roland Huß <rhuss@redhat.com> * fix(sdk/go): address principal engineer review findings - Remove dead boolCount function that would fail golangci-lint (NVIDIA#1) - Emit EventAdded for the first watch event instead of EventModified, matching k8s watch semantics (NVIDIA#7) - Add mutex locking to all mock server methods that access the shared sandboxes map, fixing latent race conditions (NVIDIA#12) - Skip HealthCheck integration test that calls an unimplemented stub (NVIDIA#13) - Scope doc.go examples: mark sections for sub-clients not yet available in this PR with "available in a future release" (NVIDIA#4) - Document Config.Timeout/RetryPolicy/Logger and WatchOptions fields as reserved for future use (NVIDIA#2, NVIDIA#6) Signed-off-by: Roland Huß <rhuss@redhat.com> * refactor(sdk/go): migrate mise config to centralized task include Move Go SDK mise configuration from standalone sdk/go/mise.toml into the project's centralized pattern: - Add Go tools (go, golangci-lint, protoc-gen-go, protoc-gen-go-grpc) to root mise.toml [tools] section - Create tasks/go.toml with all SDK tasks using go: namespace prefix and dir=sdk/go for working directory - Update sdk/go/Makefile to reference namespaced task names - Update proto:sync default path for monorepo layout Addresses review feedback from drew on PR NVIDIA#2271 regarding mise convention alignment. Signed-off-by: Roland Huß <rhuss@redhat.com> * refactor(sdk/go): remove UPSTREAM_VERSION standalone repo artifact Remove sdk/go/proto/UPSTREAM_VERSION file and its exclusion from proto:check. This was a leftover from the standalone repo prototype. In a monorepo, proto drift is detectable via git diff between sdk/go/proto/ and proto/ directly. Signed-off-by: Roland Huß <rhuss@redhat.com> * refactor(sdk/go): switch proto generation from protoc to buf Replace raw protoc invocations with buf for Go SDK proto code generation, aligning with the TS SDK approach (PR NVIDIA#2122). - Add repo-level buf.yaml declaring proto/ as the buf module with lint and breaking change detection config - Add sdk/go/buf.gen.yaml configuring buf to generate Go code directly from root proto/ (no more vendored .proto copies) - Delete vendored .proto source files from sdk/go/proto/ - Rewrite go:proto:gen and go:proto:check mise tasks to use buf - Remove go:proto:sync and go:proto:clean tasks (no longer needed) - Add proto target to sdk/go/Makefile - Add buf 1.72.0 to root mise.toml tool dependencies - Include options.proto in generation (was stripped from vendored copies) - Regenerate all .pb.go files via the new buf pipeline Signed-off-by: Roland Huß <rhuss@redhat.com> * test(sdk/go): add proto-converter field coverage detection Use protobuf reflection to enumerate all fields on key proto messages (SandboxSpec, SandboxTemplate, SandboxStatus, SandboxCondition, SandboxPolicy) and compare against explicit handled/skipped sets in the converter tests. Unhandled fields produce warnings (t.Log), not failures, so proto contributors are not forced to fix SDK converters in the same PR. Stale entries in the handled set (removed proto fields) do fail, since they indicate the converter references something that no longer exists. A follow-up CI workflow will create GitHub issues when converter drift lands on main. Signed-off-by: Roland Huß <rhuss@redhat.com> * fix(sdk/go): bump Go to 1.26 and fix errcheck lint violations The upstream go.mod now has `toolchain go1.26.4`, which requires Go 1.26 to build golangci-lint. Bump the mise.toml Go version from 1.25 to 1.26 and wrap deferred Close() calls in test helpers to satisfy errcheck. Assisted-By: 🤖 Claude Code * feat(sdk/go): add ObjectMeta fields (annotations, workspace, deletion_timestamp) Add three new proto ObjectMeta fields to Sandbox and Provider domain types: Annotations (map), Workspace (string), and DeletionTimestamp (*time.Time). Update converters in both directions, deep-copy maps at the proto/SDK boundary, and add TimeFromMillisPtr/MillisFromTimePtr helper functions. Assisted-By: 🤖 Claude Code * chore(sdk/go): regenerate proto bindings after rebase Pick up workspace fields from upstream PR NVIDIA#2445 (Wire authorization into workspace model). All request messages now include workspace parameter in the generated Go bindings. Assisted-By: 🤖 Claude Code * feat(sdk/go): add workspace scoping to all RPC interfaces Add workspace parameter to every sandbox-scoped RPC method across all interfaces (Sandbox, Exec, File, Service, SSH, TCP, Config, Policy, Provider, Profile, Refresh). The workspace string is passed as the second parameter after ctx, following the convention workspace then resource-name. Key changes: - SandboxInterface: all 10 methods gain workspace parameter - sandbox_client.go: passes Workspace field in every proto request - ListOptions: add AllWorkspaces field for cross-workspace queries - All stub interfaces updated to match new signatures - All sandbox client tests updated with "default" workspace Assisted-By: 🤖 Claude Code * chore(sdk/go): remove coverage.out from tracking Assisted-By: 🤖 Claude Code * fix(sdk/go): address review feedback from mrunalp - Add RefreshStrategyAWSStsAssumeRole to match proto enum value 6, fulfilling the "all domain types upfront" contract - Wrap context.DeadlineExceeded and context.Canceled in StatusError so IsDeadlineExceeded() and IsCancelled() helpers work correctly - Return error from mapToStruct/SandboxSpecToProto instead of silently discarding structpb.NewStruct failures on invalid template maps Signed-off-by: Roland Huss <rhuss@redhat.com> * fix(sdk/go): address remaining review items - Wire go:ci into root ci task so SDK is tested in repository CI - Fix gofmt formatting on converter files - Add goimports to mise.toml tools - Add coverage.out to .gitignore - Add Go SDK section to AGENTS.md and CONTRIBUTING.md - Add regression tests for context-error wrapping (IsDeadlineExceeded, IsCancelled) and invalid template map rejection - Remove panic from SandboxToProto, return error instead Signed-off-by: Roland Huss <rhuss@redhat.com> * fix(sdk/go): pin goimports version and update lockfile Pin goimports to 0.48.0 instead of "latest" and regenerate mise.lock to include the new entry. Signed-off-by: Roland Huss <rhuss@redhat.com> * fix(sdk/go): TLS.Insecure means skip-verify, not plaintext Align TLS.Insecure semantics with the Rust SDK: Insecure: true now uses TLS with InsecureSkipVerify (skip cert verification) instead of switching to plaintext. Only the http:// scheme triggers plaintext. This fixes token auth against dev/k3d gateways: StaticToken and RefreshableToken require transport security, which real TLS (even with InsecureSkipVerify) satisfies, but plaintext does not. For http:// + token auth (dev gateways without TLS), wrap the auth provider to override RequireTransportSecurity, matching the Rust SDK's behavior where http:// accepts any auth mode. Transport decision table (matches Rust SDK crates/openshell-sdk): http:// + any TLS config -> plaintext (TLS config ignored) https:// + Insecure: true -> TLS, skip cert verify https:// + Insecure: false -> TLS, full verification no scheme -> same as https:// Signed-off-by: Roland Huss <rhuss@redhat.com> * feat(sdk/go): add missing policy proto fields Add 6 previously silently dropped fields to the network policy types and converters, preventing security-relevant data loss on round-trip: NetworkEndpoint fields 19-23: - CredentialSigning: SigV4 re-signing mode - SigningService: AWS service name for SigV4 - SigningRegion: AWS region override for SigV4 - JsonRpcMaxBodyBytes: JSON-RPC body inspection limit - Mcp: MCP-specific policy options (new McpOptions type) L7Allow and L7DenyRule field 9: - Params: MCP params matcher map for tools/call filtering New type McpOptions with StrictToolNames and AllowAllKnownMcpMethods optional booleans matching the proto definitions. Signed-off-by: Roland Huss <rhuss@redhat.com> * fix(sdk/go): enforce coverage test and extend to policy messages Change coverage_test.go from t.Logf (silent) to t.Errorf so that unhandled proto fields fail the test immediately. Add coverage tests for NetworkEndpoint (23 fields), L7Allow (8 fields), L7DenyRule (8 fields), and McpOptions (2 fields). Any new proto field that is not in the handled set or explicitly skipped now breaks the build, closing the silent-drift gap. Signed-off-by: Roland Huss <rhuss@redhat.com> * ci(sdk/go): add Go SDK job to branch-checks workflow Add a Go SDK job to branch-checks.yml that runs mise run go:ci (lint, build, test, proto-check, docs-check) on every PR. This ensures the SDK is tested in CI, not just locally. Signed-off-by: Roland Huss <rhuss@redhat.com> * fix(sdk/go): address should-fix review items NVIDIA#6 Fix broken godoc examples: add workspace parameter to all method calls in doc.go that were broken after workspace scoping. NVIDIA#7 Add Err field to Event[T]: Watch error events now carry the underlying error instead of discarding it. NVIDIA#8 Separate Unauthenticated from PermissionDenied: add ErrorUnauthenticated code and IsUnauthenticated() helper. gRPC Unauthenticated (401) now maps to its own code instead of collapsing into PermissionDenied (403). NVIDIA#9 Add Unwrap to StatusError: replace dead Details field with Cause error field. StatusError.Unwrap() returns Cause, enabling errors.Is/As unwrapping. FromGRPCError and contextError both populate Cause. Signed-off-by: Roland Huss <rhuss@redhat.com> * ci(sdk/go): add go:format:check to CI pipeline Add gofmt format verification to go:ci. Catches unformatted Go files before they reach the PR. Fix formatting on coverage_test.go. Signed-off-by: Roland Huss <rhuss@redhat.com> * chore(sdk/go): remove Makefile in favor of mise tasks All build, lint, test, and proto-gen tasks are already defined in tasks/go.toml and invoked via mise. The Makefile was a leftover that duplicated this and raised questions in review. Signed-off-by: Roland Huß <rhuss@redhat.com> * feat(sdk/go): sync proto bindings and add credential handle support Regenerate Go proto bindings after rebase to pick up new CredentialHandle message and Provider.credential_handles and profile_workspace fields from upstream. Add domain types, converter support, and proto field coverage tests for Provider and CredentialHandle. Signed-off-by: Roland Huß <rhuss@redhat.com> * fix(sdk/go): reject plaintext auth leak and fix watch error handling Reject http:// addresses when the auth provider requires transport security instead of silently stripping the requirement. Remove the insecureAuthWrapper that overrode RequireTransportSecurity. Fix watch stream error handling: use blocking send for terminal errors so they are never silently dropped when the channel is full, and wrap mid-stream errors with converter.FromGRPCError so SDK error helpers like IsUnavailable work on watch Event.Err. Signed-off-by: Roland Huß <rhuss@redhat.com> * fix(sdk/go): address review findings from multi-agent code review - WaitReady now detects SandboxDeleting phase and returns immediately instead of polling indefinitely - Watch goroutine defers streamCancel() to prevent context leaks - Fix StopOnTerminal=false test to keep stream open (was wrong-reason pass due to stream ending, not StopOnTerminal logic) - Add EventDeleted test covering the Deleting phase branch - Add provider converter unit tests for CredentialHandle round-trip, nil handling, and empty maps Signed-off-by: Roland Huß <rhuss@redhat.com> --------- Signed-off-by: Roland Huß <rhuss@redhat.com> Signed-off-by: Roland Huss <rhuss@redhat.com>

drew force-pushed the dn/ci-tests branch 3 times, most recently from 2698555 to bda31e5Compare February 27, 2026 07:42

wip
650d1ce

drew force-pushed the dn/ci-tests branch from bda31e5 to 650d1ceCompare February 27, 2026 16:06

drew added 11 commits February 27, 2026 08:13

wip
4f482af

wip
598bf90

ci: consolidate rust and python workflows into single checks workflow
08ffe34

ci: fix container image reference in checks workflow
915f2da

ci: use hardcoded lowercase image ref for GHCR compatibility
a15ad57

ci: add mise reshim to regenerate shims after install
3c794de

ci: add cargo bin to PATH in CI image and remove unnecessary mise ins…
978bc8a
…tall steps

ci: add cargo bin to GITHUB_PATH for rust job
dd1cfdb

ci: allow manual CI image build+push via workflow_dispatch
eea2d57

ci: simplify ci-image triggers and remove rust PATH workaround
82f3e12

ci: install all tools via mise in CI image to prevent re-downloads
616bc6c

pimlock mentioned this pull request Feb 27, 2026
Proxy improvements: better handling of non-CONNECT requests and observability #42
Closed

drew added 3 commits February 27, 2026 09:44

wip
2d0df73

ci: pass MISE_GITHUB_TOKEN as build arg in CI Dockerfile
2a69e32
Declare the ARG before mise install so authenticated GitHub API requests avoid rate limits when fetching tool versions.

drew merged commit 7e7e84e into mainFeb 27, 2026
3 checks passed

drew deleted the dn/ci-tests branch February 27, 2026 18:10

This was referenced Mar 5, 2026
Unbounded WatchSandbox sandbox_id #27
Closed
feat: sandbox-initiated policy recommendations with human-in-the-loop approval #153
Closed

drew added a commit that referenced this pull request Mar 16, 2026
ci: add GitHub Actions CI workflow with lint, test, and image build (#1)
a71f2a5

mjamiv mentioned this pull request Mar 27, 2026
feat: gateway resume from existing state and persistent SSH handshake secret #487
Closed
10 tasks

johntmyers mentioned this pull request Mar 30, 2026
feat(sandbox): extend L7 credential injection — query params, Basic auth, URL paths #689
Closed

wenlingd mentioned this pull request Mar 31, 2026
Gateway fails to start on DRIVE AGX Thor / Ferrix P3960: CONFIG_VETH not enabled in DRIVE OS RT kernel #692
Closed
3 tasks

mjamiv mentioned this pull request Apr 6, 2026
deny rules in network policy schema #565
Closed
2 tasks

vnicolici mentioned this pull request Apr 11, 2026
bug: link-local / loopback entries in allowed_ips are silently accepted but never enforced #814
Closed
3 tasks

ben-alkov pushed a commit to ben-alkov/OpenShell that referenced this pull request Apr 13, 2026
Merge pull requestNVIDIA#1from maxamillion/openshell-podman
92b0b99
feat(podman): add Podman support as alternative container runtime

ben-alkov pushed a commit to ben-alkov/OpenShell that referenced this pull request Apr 13, 2026
Merge pull requestNVIDIA#1from itdove/openshell-podman-itdove
d8908f8
docs(podman): Add macOS installation guide and Podman build compatibility

johntmyers mentioned this pull request Apr 21, 2026
Enhanced Provider Management #896
Open

unco3 mentioned this pull request Apr 24, 2026
[OpenShell] 'openshell forward list' reports STATUS: dead but there is no supervisor, auto-restart, or alerting — silently unreachable dashboards #874
Open

crdion mentioned this pull request May 7, 2026
Allow sandbox supervisor to skip inner network namespace creation on hardened / immutable hosts (rootless Podman, dm-verity, overlayroot) #1194
Closed
4 tasks

johntmyers added the test-label label May 12, 2026

This was referenced May 21, 2026
feat(auth): per-sandbox authentication to gateway #1404
Merged
feat(server): declare gRPC auth (mode + scope + role) at the handler, enforce at the router #1586
Closed

rrhubenov mentioned this pull request Jun 15, 2026
refactor(openshell-sandbox): Split sandbox into process and network subcrates. #1650
Merged
6 tasks

craig-kindo mentioned this pull request Jul 7, 2026
feat: warmpool support for OpenShell sandboxes #1447
Closed

This was referenced Jul 14, 2026
fix(supervisor-network): warn on unsupported L7 access presets #2177
Merged
feat(sdk/go): add Go SDK foundation, types, and sandbox client (A) #2271
Merged

TonyLuo-NV mentioned this pull request Jul 17, 2026
bug: sandbox bind-mount ownership breaks after host reboot (unpinned rootless userns mapping, no launch-time reconcile) #2336
Open

drew mentioned this pull request Jul 17, 2026
feat(workspace): add workspace resource model with scoping, membershi… #2243
Merged
70 tasks

This was referenced Jul 19, 2026
bug(proxy): CONNECT deny response hides binary mismatch reason #2355
Closed
fix(proxy): include OPA deny reason in CONNECT 403 response #2363
Merged

mrunalp mentioned this pull request Aug 3, 2026
feat(server): support EC and EdDSA keys in OIDC JWKS validation #2593
Open
6 tasks

Navigation Menu

Uh oh!

ci: add GitHub Actions CI workflow with lint, test, and image build - #1

ci: add GitHub Actions CI workflow with lint, test, and image build#1
drew merged 15 commits into
mainfrom
dn/ci-tests

drew commented Feb 27, 2026

Uh oh!

Uh oh!

johntmyers commented May 12, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Uh oh!

Conversation

drew commented Feb 27, 2026

Summary

Uh oh!

Uh oh!

johntmyers commented May 12, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants