Point repo, image, and module paths at openrl-dev/openrl

.claude-plugin/marketplace.json

@@ -3,7 +3,7 @@
   "description": "OpenRL: a self-hosted, Kubernetes-native post-training API that serves the Tinker API.",
   "owner": {
     "name": "GKE Labs",
-    "url": "https://github.com/gke-labs/open-rl"
+    "url": "https://github.com/openrl-dev/openrl"
   },
   "plugins": [
     {

.github/workflows/build-and-push.yml

@@ -58,7 +58,7 @@ jobs:
         id: meta
         uses: docker/metadata-action@c299e40c65443455700f0fdfc63efafe5b349051 # v5.10.0
         with:
-          images: ghcr.io/gke-labs/open-rl/${{ matrix.image_name }}
+          images: ghcr.io/openrl-dev/openrl/${{ matrix.image_name }}
           # Tags every push `sha-<short>`, the default branch `latest`, and a
           # `v*` tag push `v<x.y.z>`. The literal `v` keeps the image tag equal
           # to the git tag, which the bundle pins to. The sha outranks `latest`

.github/workflows/site.yml

@@ -52,7 +52,7 @@ jobs:
   # main deploys everywhere; a fork also deploys its website/* branches, for previews
   deploy:
     needs: build
-    if: github.event_name != 'pull_request' && (github.ref == 'refs/heads/main' || github.repository_owner != 'gke-labs')
+    if: github.event_name != 'pull_request' && (github.ref == 'refs/heads/main' || github.repository_owner != 'openrl-dev')
     runs-on: ubuntu-latest
     permissions:
       pages: write

AGENTS.md

@@ -173,7 +173,7 @@ kubectl delete pods -l timeslice.io/group=samplers --ignore-not-found
 ### Kustomize Deployment & Base Manifest Best Practices (`Tip`)
 To prevent noisy container image tag diffs inside pull requests while ensuring clean deployments across Kubernetes environments, always follow this Kustomize pattern:
 1. **Deploy via Kustomize (`-k`)**: Always apply directory manifests via `kubectl apply -k <directory>` rather than `kubectl apply -f <directory>`.
-2. **Keep Base YAMLs at `:latest` or Placeholder**: In all base/rendered YAML template files (`04-api-server.yaml`, `05-worker-pod-template.yaml`, `07-accel-timeslicer-daemonset.yaml`, `09-sampler-pod-template.yaml`, `04-deployment.yaml`, etc.), keep container image tags permanently set to a static placeholder (`e.g. image: ghcr.io/gke-labs/open-rl/server:latest` or `/API server:latest`). **Never bump image tags inside these base template files.**
+2. **Keep Base YAMLs at `:latest` or Placeholder**: In all base/rendered YAML template files (`04-api-server.yaml`, `05-worker-pod-template.yaml`, `07-accel-timeslicer-daemonset.yaml`, `09-sampler-pod-template.yaml`, `04-deployment.yaml`, etc.), keep container image tags permanently set to a static placeholder (`e.g. image: ghcr.io/openrl-dev/openrl/server:latest` or `/API server:latest`). **Never bump image tags inside these base template files.**
 3. **Single Source of Truth for Version Bumps**: When releasing or bumping image versions (`e.g. 0.1.75 -> 0.1.76`), modify **only**:
    - `VERSION` (`at repository root`)
    - The `newTag:` field inside the target environment's `kustomization.yaml` (`e.g. k8s/deploy/distributed-fft-timeslice/kustomization.yaml`)

CONTRIBUTING.md

@@ -18,7 +18,7 @@ that we need help with, including:
 * Bugs in our automation scripts
 
 If anything doesn't make sense, or doesn't work when you run it, please
-[open an issue](https://github.com/gke-labs/open-rl/issues/new) and let us know.
+[open an issue](https://github.com/openrl-dev/openrl/issues/new) and let us know.
 
 Please note that this project has a [Code of Conduct](CODE_OF_CONDUCT.md) that all
 contributors are expected to follow.
@@ -36,17 +36,17 @@ We welcome many different types of contributions, including:
 * Release management
 
 Not everything happens through a GitHub pull request. If you would like to discuss an idea
-before writing code, open a [GitHub issue](https://github.com/gke-labs/open-rl/issues) and
+before writing code, open a [GitHub issue](https://github.com/openrl-dev/openrl/issues) and
 let's talk it through.
 
 <!-- TODO: OpenRL does not yet have a public developer meeting, mailing list, or chat
      channel. Once these exist, add them here and in the README. -->
 
 ## Find an Issue
 
-Issues labelled [`good first issue`](https://github.com/gke-labs/open-rl/labels/good%20first%20issue)
+Issues labelled [`good first issue`](https://github.com/openrl-dev/openrl/labels/good%20first%20issue)
 have extra context to help you make your first contribution.
-[`help wanted`](https://github.com/gke-labs/open-rl/labels/help%20wanted) issues are suitable
+[`help wanted`](https://github.com/openrl-dev/openrl/labels/help%20wanted) issues are suitable
 for anyone who isn't a core maintainer and are a good next step after your first pull request.
 
 Sometimes there won't be any issues with these labels. That's OK — there is likely still

GOVERNANCE.md

@@ -43,7 +43,7 @@ The OpenRL project and its leadership embrace the following values:
 
 ## Maintainers
 
-OpenRL Maintainers have write access to the [project GitHub repository](https://github.com/gke-labs/open-rl).
+OpenRL Maintainers have write access to the [project GitHub repository](https://github.com/openrl-dev/openrl).
 They can merge their own patches or patches from others. The current maintainers can be found
 in [MAINTAINERS.md](MAINTAINERS.md). Maintainers collectively manage the project's resources
 and contributors.

Makefile

@@ -323,7 +323,7 @@ dashboard-apply:
 # ---------------------------------------------------------------------------
 # Images published by .github/workflows/build-and-push.yml. The names must match
 # the manifests exactly or the pin silently does nothing.
-IMAGE_REPO     ?= ghcr.io/gke-labs/open-rl
+IMAGE_REPO     ?= ghcr.io/openrl-dev/openrl
 RELEASE_IMAGES ?= server api-server client scheduler
 # Gitignored; the release job uploads everything in here.
 DIST_DIR       ?= dist

README.md

@@ -1,6 +1,6 @@
 # OpenRL: self-hosted API for your RL Infrastructure
 
-[![Release](https://img.shields.io/github/v/release/gke-labs/open-rl?label=release)](https://github.com/gke-labs/open-rl/releases/latest)
+[![Release](https://img.shields.io/github/v/release/openrl-dev/openrl?label=release)](https://github.com/openrl-dev/openrl/releases/latest)
 
 OpenRL implements [Tinker](https://tinker-docs.thinkingmachines.ai/) compatible API for fine-tuning language models that you can run on your own infrastructure (machine or a kubernetes cluster). You can use the Tinker SDK to orchestrate RL training loops by writing imperative Python code directly from your local machine.
 
@@ -161,7 +161,7 @@ Participation in this project is governed by our [Code of Conduct](CODE_OF_CONDU
 <!-- TODO: Add a developer mailing list, chat channel, and public developer meeting once
      these are set up. -->
 
-- **Questions and discussion**: [GitHub Issues](https://github.com/gke-labs/open-rl/issues)
+- **Questions and discussion**: [GitHub Issues](https://github.com/openrl-dev/openrl/issues)
 - **Reporting a vulnerability**: see our [Security Policy](SECURITY.md) — please do not open a
   public issue for security reports
 - **How the project is run**: [GOVERNANCE.md](GOVERNANCE.md)

RELEASING.md

@@ -15,10 +15,10 @@ produces a release; nothing else does.
 Images, published to GHCR by `.github/workflows/build-and-push.yml` and tagged with the git tag
 verbatim:
 
-- `ghcr.io/gke-labs/open-rl/server:<tag>`
-- `ghcr.io/gke-labs/open-rl/api-server:<tag>`
-- `ghcr.io/gke-labs/open-rl/client:<tag>` (the e2e client, `src/server/Dockerfile.client`)
-- `ghcr.io/gke-labs/open-rl/scheduler:<tag>`
+- `ghcr.io/openrl-dev/openrl/server:<tag>`
+- `ghcr.io/openrl-dev/openrl/api-server:<tag>`
+- `ghcr.io/openrl-dev/openrl/client:<tag>` (the e2e client, `src/server/Dockerfile.client`)
+- `ghcr.io/openrl-dev/openrl/scheduler:<tag>`
 
 Assets, attached to the GitHub Release:
 
@@ -71,7 +71,7 @@ Every bundle has its images pinned to the release tag. The API server reports th
 4. **Verify on a clean cluster.**
 
    ```bash
-   kubectl apply --server-side -f https://github.com/gke-labs/open-rl/releases/download/v0.0.2/openrl-lora.yaml
+   kubectl apply --server-side -f https://github.com/openrl-dev/openrl/releases/download/v0.0.2/openrl-lora.yaml
    kubectl -n openrl-system get pods -o jsonpath='{..image}'
    ```
 

SECURITY.md

@@ -7,7 +7,7 @@ If you discover a security vulnerability in OpenRL, please report it responsibly
 
 To report a vulnerability, use GitHub Private Vulnerability Reporting:
 
-* **[Report a vulnerability](https://github.com/gke-labs/open-rl/security/advisories/new)**
+* **[Report a vulnerability](https://github.com/openrl-dev/openrl/security/advisories/new)**
 
 <!-- TODO: Set up a project security email (e.g. openrl-security@googlegroups.com) and
      add it here as an alternative reporting channel for reporters who cannot or prefer
@@ -49,7 +49,7 @@ When a security issue is confirmed, the OpenRL maintainers will:
 2. Assign a CVE identifier if appropriate
 3. Release a patched version
 4. Publish a security advisory via
-   [GitHub Security Advisories](https://github.com/gke-labs/open-rl/security/advisories)
+   [GitHub Security Advisories](https://github.com/openrl-dev/openrl/security/advisories)
 
 ## Security Response Team
 

docs/blog/from-mac-to-gke.md

@@ -2,7 +2,7 @@
 
 RL fine-tuning is one of the most powerful ways to specialize language models — but the infrastructure behind it has traditionally been a nightmare. You're either wrestling with GPU allocation, rewriting training scripts for different backends, or managing job lifecycles by hand.
 
-[OpenRL](https://github.com/google/open-rl) is a self-hosted, open-source API that makes this simple. Write your training loop once using the Tinker SDK, run it on your Mac to iterate fast, then point it at a GKE cluster when you're ready to scale. Same code, any backend.
+[OpenRL](https://github.com/openrl-dev/openrl) is a self-hosted, open-source API that makes this simple. Write your training loop once using the Tinker SDK, run it on your Mac to iterate fast, then point it at a GKE cluster when you're ready to scale. Same code, any backend.
 
 Let's walk through it.
 
@@ -66,7 +66,7 @@ Here's the key insight: **your client code doesn't change.** The Tinker API abst
 
 ```bash
 # Deploy to GKE
-kubectl apply --server-side -f https://github.com/gke-labs/open-rl/releases/latest/download/openrl-lora.yaml
+kubectl apply --server-side -f https://github.com/openrl-dev/openrl/releases/latest/download/openrl-lora.yaml
 
 # Port-forward to your local machine
 kubectl -n openrl-system port-forward svc/open-rl-api-server-service 8000:8000

docs/designs/007-configurable-dra-claim-pools-and-worker-scheduler.md

@@ -182,7 +182,7 @@ spec:
   # claimRef: "open-rl-trainer-gpu-1"
 
   # Container overrides (image, args, env)
-  image: "ghcr.io/gke-labs/open-rl/server:latest"
+  image: "ghcr.io/openrl-dev/openrl/server:latest"
 
 status:
   phase: "Scheduled" # Pending | Scheduled | Running | Failed | Terminated

docs/designs/008-single-node-shared-training-sampling.md

@@ -210,7 +210,7 @@ patches:
             restartPolicy: OnFailure
             containers:
             - name: trainer-worker
-              image: ghcr.io/gke-labs/open-rl/server:latest
+              image: ghcr.io/openrl-dev/openrl/server:latest
               imagePullPolicy: IfNotPresent
               command: ["uv", "run", "python", "-m", "server.training_requests_processor"]
               env:

docs/designs/014-openrl-website.md

@@ -229,10 +229,10 @@ uses for keywords, strings, comments and highlighted lines.
   openrl.dev.      AAAA  2606:50c0:8001::153
   openrl.dev.      AAAA  2606:50c0:8002::153
   openrl.dev.      AAAA  2606:50c0:8003::153
-  www.openrl.dev.  CNAME gke-labs.github.io.
+  www.openrl.dev.  CNAME openrl-dev.github.io.
   ```
 
-  GitHub also asks for a `_github-pages-challenge-gke-labs` TXT record
+  GitHub also asks for a `_github-pages-challenge-openrl-dev` TXT record
   to verify the domain for the organization; the value comes from the
   organization's Pages settings.
 - openrl.io: GitHub Pages serves one custom domain per site, so the
@@ -284,7 +284,7 @@ git worktree add ../open-rl-site -b website/scaffold upstream/main
 cd ../open-rl-site
 hugo server -s site -D
 git push -u origin website/scaffold
-gh pr create --repo gke-labs/open-rl --base main --head droot:website/scaffold
+gh pr create --repo openrl-dev/openrl --base main --head droot:website/scaffold
 
 # when the PR has merged
 cd ../open-rl && git worktree remove ../open-rl-site
@@ -319,7 +319,7 @@ if the reviewers of those workflows agree.
 1. openrl.dev as canonical, and where the DNS for each domain is
    hosted.
 2. Hugo and GitHub Pages, or a different pair.
-3. Who has admin on gke-labs/open-rl for the Pages settings and the
+3. Who has admin on openrl-dev/openrl for the Pages settings and the
    organization-level domain verification.
 4. Which self-hosted sans. Candidates with an Avenir-like feel under
    the Open Font License: Figtree, Nunito Sans, Outfit.

docs/designs/016-agent-friendly-openrl.md

@@ -135,8 +135,8 @@ version OpenRL supports.
     "settings_howto": "Pass as user_metadata on create_session or create_model, or as TINKER_TAGS=openrl.key=value,...",
     "limits": {"vllm_max_model_len": 8192, "train_token_budget": 65536,
                "session_ttl_seconds": 120, "session_reap_interval_seconds": 30},
-    "skill_url": "https://github.com/gke-labs/open-rl/releases/download/v0.0.6/openrl-skill.tar.gz",
-    "getting_started_url": "https://github.com/gke-labs/open-rl/blob/main/docs/getting-started.md"
+    "skill_url": "https://github.com/openrl-dev/openrl/releases/download/v0.0.6/openrl-skill.tar.gz",
+    "getting_started_url": "https://github.com/openrl-dev/openrl/blob/main/docs/getting-started.md"
   }
 }
 ```
@@ -373,12 +373,12 @@ the "verified on" stamp set to the release version, listed in
 canonical; the asset is a copy, so a clone at the tag and the asset are
 the same thing. `get_server_capabilities` reports
 `openrl.skill_url` for the running version
-(`https://github.com/gke-labs/open-rl/releases/download/<version>/openrl-skill.tar.gz`),
+(`https://github.com/openrl-dev/openrl/releases/download/<version>/openrl-skill.tar.gz`),
 so an agent that starts from an endpoint finds the matching reference in
 one call. The tarball is the primary install path, since it works for
 every agent, for scripted installs, and for air-gapped clusters. Claude
 Code users get a second path, a `.claude-plugin/marketplace.json` so
-`/plugin marketplace add gke-labs/open-rl` works and can pin a tag; other
+`/plugin marketplace add openrl-dev/openrl` works and can pin a tag; other
 agents' package mechanisms can be added the same way, beside the tarball
 rather than instead of it. Installing is still a manual step for every agent, so the
 getting-started guide gets one install command right after

docs/getting-started.md

@@ -127,7 +127,7 @@ Apply the release. Server-side apply is required because the Workload CRD is
 larger than the client-side apply limit:
 
 ```bash
-kubectl apply --server-side -f https://github.com/gke-labs/open-rl/releases/latest/download/openrl-lora.yaml
+kubectl apply --server-side -f https://github.com/openrl-dev/openrl/releases/latest/download/openrl-lora.yaml
 ```
 
 This installs the API server, the scheduler, Redis, and a shared volume into
@@ -142,7 +142,7 @@ runs LoRA workers as above and adds full fine-tuning workers, which can take
 turns on a shared GPU. Apply it instead of the LoRA release:
 
 ```bash
-kubectl apply --server-side -f https://github.com/gke-labs/open-rl/releases/latest/download/openrl-fft.yaml
+kubectl apply --server-side -f https://github.com/openrl-dev/openrl/releases/latest/download/openrl-fft.yaml
 ```
 
 It adds two DaemonSets that coordinate GPU sharing on each GPU node, the
@@ -201,8 +201,8 @@ configured default.
 ## 3. Get the examples
 
 ```bash
-git clone https://github.com/gke-labs/open-rl.git
-cd open-rl
+git clone https://github.com/openrl-dev/openrl.git
+cd openrl
 uv --project examples sync
 ```
 
@@ -238,7 +238,7 @@ How to install them depends on the agent:
   directly; each one links its references by relative path.
 - **Claude Code:** both repositories publish a plugin marketplace.
   ```
-  /plugin marketplace add gke-labs/open-rl
+  /plugin marketplace add openrl-dev/openrl
   /plugin install openrl@open-rl
   /plugin marketplace add thinking-machines-lab/tinker-cookbook
   /plugin install tinker@tinker-skills
@@ -354,7 +354,7 @@ still running to release their GPUs, then everything the release created. This
 also deletes the shared volume and its data:
 
 ```bash
-kubectl delete -f https://github.com/gke-labs/open-rl/releases/latest/download/openrl-lora.yaml
+kubectl delete -f https://github.com/openrl-dev/openrl/releases/latest/download/openrl-lora.yaml
 ```
 
 Use `openrl-fft.yaml` if you applied the FFT release. If you created the GKE