## Thinking Path > - Paperclip is the open source app people use to manage AI agents for work > - Sandbox providers are the seam that lets agent runs execute in isolated environments; today the only first-party remote provider is Daytona, a hosted third-party service > - Self-hosters running Paperclip on their own infrastructure (often Kubernetes already) have no first-party way to run agent sandboxes on a cluster they control > - That gap matters for teams with data-residency, sovereignty, or cost constraints who cannot or will not send workloads to a hosted sandbox service > - This pull request adds a Kubernetes sandbox-provider plugin as a standalone, workspace-excluded package: it implements every SandboxProvider hook the Daytona provider does, on infrastructure the operator owns > - The benefit is that any Paperclip deployment with a Kubernetes cluster gets multi-tenant, network-isolated, quota-bounded agent sandboxes with zero new external dependencies ## Linked Issues or Issue Description No existing issue. Following the feature template: - **Problem:** Paperclip's remote sandbox execution requires a hosted third-party provider. Self-hosters cannot run agent sandboxes on their own Kubernetes clusters with a first-party provider. - **Proposed solution:** A `@paperclipai/plugin-kubernetes` sandbox-provider plugin with two backends: long-lived sandboxes via the [kubernetes-sigs/agent-sandbox](https://github.com/kubernetes-sigs/agent-sandbox) CRD (multi-command exec, adapter-install pattern) and one-shot `batch/v1` Jobs (stable APIs only, no extra controllers). - **Alternatives considered:** Driving kubectl from a generic shell provider (no lifecycle/lease semantics), or requiring a hosted provider (exactly the constraint this removes). ## What Changed This is **stage 1 of 3** of a staged contribution (direction agreed with maintainers): the plugin package alone. Stage 2 (server integration: lease params, provider registration) and stage 3 (agent runtime images + CI) are companion PRs that will be cross-linked from a comment here. - New package `packages/plugins/sandbox-providers/kubernetes` (workspace-excluded, like the path already carved out in `pnpm-workspace.yaml`): src, unit + kind integration tests, operator prerequisite manifests, README, smoke-test guide - Implements the full SandboxProvider hook surface the Daytona provider implements: `validateConfig`, `probe`, `acquireLease`, `resumeLease`, `releaseLease`, `destroyLease`, `realizeWorkspace`, `execute` - Two backends: `sandbox-cr` (default; long-lived pod via the agent-sandbox `Sandbox` CR, supports multi-command exec) and `job` (one-shot `batch/v1` Job; nothing beyond k8s 1.27+ required) - Per-run adapter resolution: one environment serves mixed harnesses; the per-run `adapterType` hint is read through a local optional type extension, so the plugin typechecks and builds against the current plugin SDK and simply falls back to the environment's configured default adapter until stage 2 lands - Exec-env wrapping: the Kubernetes exec API carries no environment, so commands are wrapped to receive the run's env - Fast-upload interception for workspace realization, scoped per lease - Per-tenant isolation: derived namespace per company, RBAC, ResourceQuota, restricted-PSS pod security (runAsNonRoot, drop ALL, seccomp RuntimeDefault, no SA token automount) - Network egress policy in two flavors: native `NetworkPolicy` and `CiliumNetworkPolicy` (FQDN allowlists) - Image allowlist with glob matching, registry override, and per-run image override validation - Per-run Kubernetes Secrets carrying agent credentials, ownerRef'd to the Job or Sandbox CR for cascade GC ## Verification - Standalone build, exactly as the README documents: ```bash cd packages/plugins/sandbox-providers/kubernetes pnpm install --ignore-workspace pnpm test # 147 unit tests, 17 files, all green pnpm typecheck # clean against the in-repo plugin SDK on master pnpm build # dist/ emitted, manifest + worker entrypoints present ``` - A kind-cluster end-to-end integration test is included (`RUN_K8S_INTEGRATION_TESTS=1 pnpm test test/integration/end-to-end-run.test.ts`) - Beyond CI: this provider has been verified in a production multi-tenant deployment against five harnesses (opencode, pi, codex, gemini, claude code) with real billed runs ## Risks - **Zero behavior change for any existing deployment.** The package is workspace-excluded; nothing in the server imports or loads it until stage 2's integration lands. No existing code paths are touched. - The default `sandbox-cr` backend depends on an alpha CRD (`agents.x-k8s.io/v1alpha1`); the README flags this and the `job` backend uses only stable APIs as a fallback. - Risk surface is confined to deployments that explicitly install and configure the plugin. - The default runtime images (`ghcr.io/paperclipai/agent-runtime-*`) are published by the stage 3 companion PR (#7934); until that lands, deployments must point `runtimeImage` at their own images. ## Model Used Claude Opus 4.8 (1M context), extended thinking, with tool use (Claude Code). ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [ ] If this change affects the UI, I have included before/after screenshots (no UI changes) - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [ ] All Paperclip CI gates are green (pending this push) - [ ] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
4.0 KiB
Manual smoke test — @paperclipai/plugin-kubernetes
Manual sanity check that the plugin works end-to-end against a real paperclip-server instance and a real Kubernetes cluster (kind for local dev). Future work may automate this in CI.
Prerequisites
- A running kind cluster:
kind create cluster --name paperclip kubectl --context kind-paperclip get nodesreturns a node inReadystate.
Steps
1. Build the plugin
cd packages/plugins/sandbox-providers/kubernetes
pnpm install --ignore-workspace
pnpm build
Expected: dist/ populated with compiled .js and .d.ts files. No errors.
2. Start paperclip-server in dev mode
In a separate terminal:
cd /path/to/paperclip
export PAPERCLIP_HOME=/tmp/paperclip-smoke
export PAPERCLIP_INSTANCE_ID=smoke
export PAPERCLIP_DEPLOYMENT_MODE=local_trusted
pnpm --filter @paperclipai/server dev
Wait for Server listening on 127.0.0.1:3100.
3. Install the plugin via the CLI
pnpm paperclipai plugin install \
--local /path/to/paperclip/packages/plugins/sandbox-providers/kubernetes \
--api-base http://127.0.0.1:3100
Expected: ✓ Installed paperclip.kubernetes-sandbox-provider v0.1.0 (ready).
4. Create a company and a kubernetes sandbox environment
CO_ID=$(curl -s -X POST -H "Content-Type: application/json" \
-d '{"name":"SmokeCo"}' \
http://127.0.0.1:3100/api/companies | jq -r '.id')
KUBECONFIG_CONTENT=$(cat ~/.kube/config | jq -Rs .)
curl -s -X POST -H "Content-Type: application/json" \
-d "{
\"name\": \"k8s-sandbox\",
\"driver\": \"sandbox\",
\"config\": {
\"provider\": \"kubernetes\",
\"kubeconfig\": $KUBECONFIG_CONTENT,
\"companySlug\": \"smoke\",
\"adapterType\": \"claude_local\",
\"imageAllowList\": [\"ghcr.io/paperclipai/agent-runtime-claude:v1\"]
}
}" \
http://127.0.0.1:3100/api/companies/$CO_ID/environments | jq
Expected: HTTP 201 with the new environment row.
5. Probe the environment
ENV_ID=$(curl -s http://127.0.0.1:3100/api/companies/$CO_ID/environments | jq -r '.[0].id')
curl -s -X POST -d '{}' -H "Content-Type: application/json" \
http://127.0.0.1:3100/api/environments/$ENV_ID/probe | jq
Expected: {"ok": true, ...} with a summary mentioning the tenant namespace
(paperclip-smoke). On first probe the namespace may not yet exist —
the plugin treats a 404 on listNamespacedPod as a successful reachability
check.
6. Trigger an agent run
Use the UI or the API to dispatch a run against the k8s-sandbox environment.
The plugin's onEnvironmentAcquireLease will:
ensureTenant— provision thepaperclip-smokenamespace, SA, Role, RoleBinding, ResourceQuota, LimitRange, NetworkPoliciesbuildJobManifest— render the security-hardened Job manifestcreateJob— submit tobatch/v1createPerRunSecret— owned by the Job for cascade-delete
7. Verify the tenant resources
kubectl --context kind-paperclip get namespace paperclip-smoke
kubectl --context kind-paperclip get all,networkpolicy,resourcequota,limitrange,sa,role,rolebinding -n paperclip-smoke
Expected:
- Namespace
paperclip-smokeexists with PSS labels (pod-security.kubernetes.io/enforce=restricted) - ServiceAccount
paperclip-tenant-sa - Role
paperclip-tenant-role, RoleBindingpaperclip-tenant-rb - ResourceQuota
paperclip-quota, LimitRangepaperclip-limits - NetworkPolicies
paperclip-deny-all+paperclip-egress-allow - Job
pc-{ulid}and its child Pod - Secret
pc-{ulid}-envwithownerReferencespointing at the Job
8. Tear down
kubectl --context kind-paperclip delete namespace paperclip-smoke
kill %1 # paperclip-server
9. Document the result
In the PR description (or appended to this file as a dated section), record:
- Date + git SHA
kubectl versionserver version- Output of
kubectl get all -n paperclip-smokeafter step 6 - Probe response from step 5
- Time-to-acquire-lease (target: <30s on kind for a cold tenant)