Files
PaperClipAI/packages/plugins/sdk/src/define-plugin.ts
T
b247bf7150 feat(runtime): opt-in sandbox file-sync lifecycle hooks (API + provider docs) (#10013)
## Thinking Path

> - Paperclip is an open-source AI-agent management platform; agents run
tasks inside sandboxed environments (Daytona, Kubernetes, E2B, etc.)
> - The control-plane ↔ sandbox file-transfer path flows through the
`environmentExecute` seam in `protocol.ts` — the only verb available to
plugins — which forces a base64-over-exec chunked loop for every file
move: workspace files, assets, Codex home sync
> - This transport is correct and safe, but it bypasses provider-native
bulk/streaming APIs (Daytona `uploadFiles`, K8s `FastUploadInterceptor`
/ volume mounts), leaving significant throughput on the table for large
workspaces
> - The right fix is an opt-in seam extension: providers with faster
native transfer declare two optional verbs; providers that do not opt in
stay on the existing fallback with zero code or behavior change required
> - This PR adds the first layer of that extension — two optional verbs
(`environmentSyncIn` / `environmentSyncOut`) in the plugin SDK, the
runtime plumbing to prefer the native path for the two clean
destroy-then-replace cases, and a doc for the contract
> - The core correctness invariant is byte-identical fallback: if no
provider opts in, execution is exactly what ships today;
`assertSyncOperationsConfined` enforces host-side path confinement for
providers that do opt in
> - No provider advertises the verbs yet → zero production behavior
change; future PRs wire up Daytona and K8s providers against this
contract

## Linked Issues or Issue Description

No public GitHub issue exists for this feature. Description follows the
`feature_request` issue template:

**Subsystem affected:**
packages/plugins — plugin system; packages/adapter-utils — adapter
runtime; server/ — EnvironmentRuntimeService

**Problem or motivation:**
Sandbox file transfers currently always use a base64-over-exec chunked
loop regardless of what the underlying provider supports. For workspaces
larger than a few MB this becomes the dominant wall-clock cost of every
sandbox run, and it bypasses bulk/stream APIs that providers like
Daytona already expose natively.

**Proposed solution:**
Add two optional, opt-in plugin hooks — `onEnvironmentSyncIn` /
`onEnvironmentSyncOut` — to the plugin SDK. When a provider defines both
hooks and both are advertised via the existing `supportedMethods`
negotiation, the runtime prefers the native path for the two clean
destroy-then-replace transfer cases; all other cases fall back to the
existing byte-identical base64 transport.

**Alternatives considered:**
An unconditional verb would require every provider to implement or stub
the verb. The opt-in / `METHOD_NOT_IMPLEMENTED` pattern (already used by
`environmentExecute`) preserves backward compatibility with zero
provider changes required.

**Roadmap alignment:**
Consistent with the ✅ "Cloud / Sandbox agents" and ✅ "Plugin system"
milestones; extends the plugin seam rather than adding
control-plane-level logic.

**Additional context:**
Searched open pull requests and issues for duplicate sandbox file-sync /
native-transfer work; none found.

## What Changed

- **`packages/plugins/sdk`**
- `protocol.ts`: two new optional `HostToWorkerMethods` —
`environmentSyncIn` / `environmentSyncOut` — plus generic
`SyncOperation`, `SyncFileMapping`, and `SyncOutcome` types
- `define-plugin.ts`: optional `onEnvironmentSyncIn` /
`onEnvironmentSyncOut` fields on `PluginDefinition`; worker advertises
each verb only when its hook is defined (else `METHOD_NOT_IMPLEMENTED`,
mirroring `environmentExecute`)
  - `worker-rpc-host.ts`: route new verbs to plugin hooks
  - `index.ts`: re-export new public types
- **`packages/adapter-utils`**
- `command-managed-runtime.ts`: expose optional `syncIn` / `syncOut` on
`CommandManagedRuntimeRunner` (available only when both verbs are
advertised); add `assertSyncOperationsConfined` host-side
path-confinement guard
- `sandbox-managed-runtime.ts`: `SandboxManagedRuntimeClient` gains
optional `syncIn` / `syncOut`; orchestrator prefers native path for
default-provision asset inbound and workspace-download-into-fresh-dir
outbound; all other paths keep the existing base64 fallback
- `sandbox-file-sync.test.ts` (new): 234-line characterization suite —
native-opt-in branch, fallback branch, `assertSyncOperationsConfined`
escape-path rejection, `followSymlinks` → tar `-h`
- `command-managed-runtime.test.ts`: negotiation + native-sync +
confinement tests
- **`server/src/services/environment-runtime.ts`**:
`EnvironmentRuntimeService` delegates to `syncIn` / `syncOut`, gated on
advertised support
- **`server/src/services/environment-execution-target.ts`**: minor
typing fix alongside the new verbs
- **`doc/plugins/SANDBOX_FILE_SYNC_HOOKS.md`** (new): documents the full
contract — opt-in / no-op guarantee, operation ordering,
provider-may-tar, atomicity, `followSymlinks`, secret modes (0600, no
window), path confinement, `operationId` opacity, resource bounds,
shell-quoting

## Verification

```bash
# SDK suite
pnpm --filter packages/plugins/sdk test

# Adapter-utils suite (includes new sandbox-file-sync characterization tests)
pnpm --filter packages/adapter-utils test
# Expected: 255 pass / 4 skip

# Type-check across affected packages
pnpm --filter packages/plugins/sdk typecheck
pnpm --filter packages/adapter-utils typecheck
# Server changed-file spot check:
cd server && npx tsc --noEmit --skipLibCheck 2>&1 | grep -E "environment-(runtime|execution-target)" | head -20
```

Key behavioral invariant to spot-check: with no provider opting in (the
current state), run any sandbox task and confirm file-transfer behavior
is byte-for-byte identical to what the pre-PR code produces. The
characterization tests assert this at the unit level.

## Risks

- **Zero production risk today**: no provider advertises
`environmentSyncIn` / `environmentSyncOut`, so the new code paths are
unreachable in production; all real traffic stays on the existing base64
fallback
- **Path confinement**: `assertSyncOperationsConfined` rejects any
`targetPath` that escapes the declared root — this is the primary
security boundary for future providers. The test suite covers
escape-path rejection
- **Atomicity**: the contract delegates atomicity to providers; the doc
explicitly calls out that directory-level ops are not guaranteed atomic
- **Secret transport**: credential assets (e.g., Codex `auth.json`,
directory mappings) continue to use the existing tar path — they do not
go through the new verbs in any current provider

> For core feature work, check [`ROADMAP.md`](ROADMAP.md) first and
discuss it in `#dev` before opening the PR. Feature PRs that overlap
with planned core work may need to be redirected — check the roadmap
first. See `CONTRIBUTING.md`.

## Model Used

Provider: Anthropic  
Model: `claude-sonnet-4-6` (Claude Sonnet 4.6)  
Context window: 200 K tokens  
Capabilities: extended tool use, multi-file code generation, agentic
reasoning via the Paperclip agent framework

## 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 not referenced internal/instance-local Paperclip issues or
links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip`
URLs)
- [x] My branch name describes the change (e.g. `docs/...`, `fix/...`)
and contains no internal Paperclip ticket id or instance-derived details
- [x] I have run tests locally and they pass
- [x] I have added or updated tests where applicable
- [x] I have updated relevant documentation to reflect my changes
- [x] I have considered and documented any risks above
- [x] All Paperclip CI gates are green
- [x] 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: Harold Kim <harold@paperclip.ing>
Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-07-22 10:08:03 -07:00

443 lines
15 KiB
TypeScript

/**
* `definePlugin` — the top-level helper for authoring a Paperclip plugin.
*
* Plugin authors call `definePlugin()` and export the result as the default
* export from their worker entrypoint. The host imports the worker module,
* calls `setup()` with a `PluginContext`, and from that point the plugin
* responds to events, jobs, webhooks, and UI requests through the context.
*
* @see PLUGIN_SPEC.md §14.1 — Example SDK Shape
*
* @example
* ```ts
* // dist/worker.ts
* import { definePlugin } from "@paperclipai/plugin-sdk";
*
* export default definePlugin({
* async setup(ctx) {
* ctx.logger.info("Linear sync plugin starting");
*
* // Subscribe to events
* ctx.events.on("issue.created", async (event) => {
* const companyId = event.companyId;
* const config = await ctx.config.get(companyId);
* const apiKey = await ctx.secrets.resolve(config.apiKeyRef, { companyId, configPath: "apiKeyRef" });
* await ctx.http.fetch(`https://api.linear.app/...`, {
* method: "POST",
* headers: { Authorization: `Bearer ${apiKey}` },
* body: JSON.stringify({ title: event.payload.title }),
* });
* });
*
* // Register a job handler
* ctx.jobs.register("full-sync", async (job) => {
* ctx.logger.info("Running full-sync job", { runId: job.runId });
* // ... sync logic
* });
*
* // Register data for the UI
* ctx.data.register("sync-health", async ({ companyId }) => {
* const state = await ctx.state.get({
* scopeKind: "company",
* scopeId: String(companyId),
* stateKey: "last-sync",
* });
* return { lastSync: state };
* });
* },
* });
* ```
*/
import type { PluginContext } from "./types.js";
import type {
PluginEnvironmentAcquireLeaseParams,
PluginEnvironmentDestroyLeaseParams,
PluginEnvironmentExecuteParams,
PluginEnvironmentExecuteResult,
PluginEnvironmentSyncInParams,
PluginEnvironmentSyncOutParams,
PluginEnvironmentSyncResult,
PluginEnvironmentStartInteractiveSetupParams,
PluginEnvironmentInteractiveSetupSession,
PluginEnvironmentGetInteractiveSetupParams,
PluginEnvironmentCaptureTemplateParams,
PluginEnvironmentCaptureTemplateResult,
PluginEnvironmentCancelInteractiveSetupParams,
PluginEnvironmentCancelInteractiveSetupResult,
PluginEnvironmentDeleteTemplateParams,
PluginEnvironmentDeleteTemplateResult,
PluginEnvironmentLease,
PluginEnvironmentProbeParams,
PluginEnvironmentProbeResult,
PluginEnvironmentRealizeWorkspaceParams,
PluginEnvironmentRealizeWorkspaceResult,
PluginEnvironmentReleaseLeaseParams,
PluginEnvironmentResumeLeaseParams,
PluginEnvironmentValidateConfigParams,
PluginEnvironmentValidationResult,
DetectExternalObjectsParams,
DetectExternalObjectsResult,
ResolveExternalObjectParams,
PluginExternalObjectResolveResult,
RefreshExternalObjectsParams,
RefreshExternalObjectsResult,
} from "./protocol.js";
// ---------------------------------------------------------------------------
// Health check result
// ---------------------------------------------------------------------------
/**
* Optional plugin-reported diagnostics returned from the `health()` RPC method.
*
* @see PLUGIN_SPEC.md §13.2 — `health`
*/
export interface PluginHealthDiagnostics {
/** Machine-readable status: `"ok"` | `"degraded"` | `"error"`. */
status: "ok" | "degraded" | "error";
/** Human-readable description of the current health state. */
message?: string;
/** Plugin-reported key-value diagnostics (e.g. connection status, queue depth). */
details?: Record<string, unknown>;
}
// ---------------------------------------------------------------------------
// Config validation result
// ---------------------------------------------------------------------------
/**
* Result returned from the `validateConfig()` RPC method.
*
* @see PLUGIN_SPEC.md §13.3 — `validateConfig`
*/
export interface PluginConfigValidationResult {
/** Whether the config is valid. */
ok: boolean;
/** Non-fatal warnings about the config. */
warnings?: string[];
/** Validation errors (populated when `ok` is `false`). */
errors?: string[];
}
// ---------------------------------------------------------------------------
// Webhook handler input
// ---------------------------------------------------------------------------
/**
* Input received by the plugin worker's `handleWebhook` handler.
*
* @see PLUGIN_SPEC.md §13.7 — `handleWebhook`
*/
export interface PluginWebhookInput {
/** Endpoint key matching the manifest declaration. */
endpointKey: string;
/** Inbound request headers. */
headers: Record<string, string | string[]>;
/** Raw request body as a UTF-8 string. */
rawBody: string;
/** Parsed JSON body (if applicable and parseable). */
parsedBody?: unknown;
/** Unique request identifier for idempotency checks. */
requestId: string;
}
export interface PluginApiRequestInput {
routeKey: string;
method: string;
path: string;
params: Record<string, string>;
query: Record<string, string | string[]>;
body: unknown;
actor: {
actorType: "user" | "agent";
actorId: string;
agentId?: string | null;
userId?: string | null;
runId?: string | null;
};
companyId: string;
headers: Record<string, string>;
}
export interface PluginApiResponse {
status?: number;
headers?: Record<string, string>;
body?: unknown;
}
// ---------------------------------------------------------------------------
// Plugin definition
// ---------------------------------------------------------------------------
/**
* The plugin definition shape passed to `definePlugin()`.
*
* The only required field is `setup`, which receives the `PluginContext` and
* is where the plugin registers its handlers (events, jobs, data, actions,
* tools, etc.).
*
* All other lifecycle hooks are optional. If a hook is not implemented the
* host applies default behaviour (e.g. restarting the worker on config change
* instead of calling `onConfigChanged`).
*
* @see PLUGIN_SPEC.md §13 — Host-Worker Protocol
*/
export interface PluginDefinition {
/**
* Called once when the plugin worker starts up, after `initialize` completes.
*
* This is where the plugin registers all its handlers: event subscriptions,
* job handlers, data/action handlers, and tool registrations. Registration
* must be synchronous after `setup` resolves — do not register handlers
* inside async callbacks that may resolve after `setup` returns.
*
* @param ctx - The full plugin context provided by the host
*/
setup(ctx: PluginContext): Promise<void>;
/**
* Called when the host wants to know if the plugin is healthy.
*
* The host polls this on a regular interval and surfaces the result in the
* plugin health dashboard. If not implemented, the host infers health from
* worker process liveness.
*
* @see PLUGIN_SPEC.md §13.2 — `health`
*/
onHealth?(): Promise<PluginHealthDiagnostics>;
/**
* Called when the operator updates this plugin's company-scoped configuration
* at runtime, without restarting the worker.
*
* If not implemented, the host restarts the worker to apply the new config.
*
* @param newConfig - The newly resolved configuration
* @see PLUGIN_SPEC.md §13.4 — `configChanged`
*/
onConfigChanged?(newConfig: Record<string, unknown>): Promise<void>;
/**
* Called when the host is about to shut down the plugin worker.
*
* The worker has at most 10 seconds (configurable via plugin config) to
* finish in-flight work and resolve this promise. After the deadline the
* host sends SIGTERM, then SIGKILL.
*
* @see PLUGIN_SPEC.md §12.5 — Graceful Shutdown Policy
*/
onShutdown?(): Promise<void>;
/**
* Called to validate the current plugin configuration.
*
* The host calls this:
* - after the plugin starts (to surface config errors immediately)
* - after the operator saves a new config (to validate before persisting)
* - via the "Test Connection" button in the settings UI
*
* @param config - The configuration to validate
* @see PLUGIN_SPEC.md §13.3 — `validateConfig`
*/
onValidateConfig?(config: Record<string, unknown>): Promise<PluginConfigValidationResult>;
/**
* Called to handle an inbound webhook delivery.
*
* The host routes `POST /api/plugins/:pluginId/webhooks/:endpointKey` to
* this handler. The plugin is responsible for signature verification using
* a resolved secret ref.
*
* If not implemented but webhooks are declared in the manifest, the host
* returns HTTP 501 for webhook deliveries.
*
* @param input - Webhook delivery metadata and payload
* @see PLUGIN_SPEC.md §13.7 — `handleWebhook`
*/
onWebhook?(input: PluginWebhookInput): Promise<void>;
/**
* Called for manifest-declared scoped JSON API routes under
* `/api/plugins/:pluginId/api/*` after the host has enforced auth, company
* access, capabilities, and checkout policy.
*/
onApiRequest?(input: PluginApiRequestInput): Promise<PluginApiResponse>;
/**
* Called when Paperclip scans issue/comment/document content and asks this
* plugin whether any sanitized URL candidates belong to its external object
* providers. The host has already stripped URL userinfo, query strings, and
* fragments unless provider-safe identity components were explicitly hashed.
*
* Requires `external.objects.detect`.
*/
onDetectExternalObjects?(
params: DetectExternalObjectsParams,
): Promise<DetectExternalObjectsResult>;
/**
* Called when Paperclip needs the current normalized status for one external
* object owned by a manifest-declared provider.
*
* Requires `external.objects.read`.
*/
onResolveExternalObject?(
params: ResolveExternalObjectParams,
): Promise<PluginExternalObjectResolveResult>;
/**
* Optional batch resolver used by providers that can refresh many objects
* more efficiently than individual `onResolveExternalObject` calls.
*
* Requires `external.objects.refresh`.
*/
onRefreshExternalObjects?(
params: RefreshExternalObjectsParams,
): Promise<RefreshExternalObjectsResult>;
/**
* Called to validate provider-specific configuration for a plugin-hosted
* environment driver.
*/
onEnvironmentValidateConfig?(
params: PluginEnvironmentValidateConfigParams,
): Promise<PluginEnvironmentValidationResult>;
/** Called to test reachability or readiness of a plugin-hosted environment. */
onEnvironmentProbe?(
params: PluginEnvironmentProbeParams,
): Promise<PluginEnvironmentProbeResult>;
/** Called before a run starts to acquire a provider lease. */
onEnvironmentAcquireLease?(
params: PluginEnvironmentAcquireLeaseParams,
): Promise<PluginEnvironmentLease>;
/** Called to reconnect to a previously acquired provider lease. */
onEnvironmentResumeLease?(
params: PluginEnvironmentResumeLeaseParams,
): Promise<PluginEnvironmentLease>;
/** Called when a run finishes and the provider lease can be released. */
onEnvironmentReleaseLease?(
params: PluginEnvironmentReleaseLeaseParams,
): Promise<void>;
/** Called when the host needs to force-destroy provider state. */
onEnvironmentDestroyLease?(
params: PluginEnvironmentDestroyLeaseParams,
): Promise<void>;
/** Called to materialize the run workspace inside the provider lease. */
onEnvironmentRealizeWorkspace?(
params: PluginEnvironmentRealizeWorkspaceParams,
): Promise<PluginEnvironmentRealizeWorkspaceResult>;
/** Called to execute a command inside the provider lease. */
onEnvironmentExecute?(
params: PluginEnvironmentExecuteParams,
): Promise<PluginEnvironmentExecuteResult>;
/**
* Optional, opt-in: called before execution to place host files/directories at
* target sandbox paths using a provider-native transport instead of the default
* base64-over-exec fallback. Defining this hook (together with
* `onEnvironmentSyncOut`) advertises `environmentSyncIn`; leaving it undefined
* keeps the byte-identical fallback. See `doc/plugins/SANDBOX_FILE_SYNC_HOOKS.md`.
*/
onEnvironmentSyncIn?(
params: PluginEnvironmentSyncInParams,
): Promise<PluginEnvironmentSyncResult>;
/**
* Optional, opt-in: called after execution to copy sandbox files/directories
* back to target host paths using a provider-native transport. Defining this
* hook (together with `onEnvironmentSyncIn`) advertises `environmentSyncOut`.
* See `doc/plugins/SANDBOX_FILE_SYNC_HOOKS.md`.
*/
onEnvironmentSyncOut?(
params: PluginEnvironmentSyncOutParams,
): Promise<PluginEnvironmentSyncResult>;
/** Called to start an interactive setup sandbox and return redacted connection metadata. */
onEnvironmentStartInteractiveSetup?(
params: PluginEnvironmentStartInteractiveSetupParams,
): Promise<PluginEnvironmentInteractiveSetupSession>;
/** Called to read setup status and, when authorized, a one-time connection payload. */
onEnvironmentGetInteractiveSetup?(
params: PluginEnvironmentGetInteractiveSetupParams,
): Promise<PluginEnvironmentInteractiveSetupSession>;
/** Called to capture a reusable provider template from a live setup sandbox. */
onEnvironmentCaptureTemplate?(
params: PluginEnvironmentCaptureTemplateParams,
): Promise<PluginEnvironmentCaptureTemplateResult>;
/** Called to cancel and clean up a setup sandbox without promoting a template. */
onEnvironmentCancelInteractiveSetup?(
params: PluginEnvironmentCancelInteractiveSetupParams,
): Promise<PluginEnvironmentCancelInteractiveSetupResult>;
/** Called for optional best-effort cleanup of a captured provider template. */
onEnvironmentDeleteTemplate?(
params: PluginEnvironmentDeleteTemplateParams,
): Promise<PluginEnvironmentDeleteTemplateResult>;
}
// ---------------------------------------------------------------------------
// PaperclipPlugin — the sealed object returned by definePlugin()
// ---------------------------------------------------------------------------
/**
* The sealed plugin object returned by `definePlugin()`.
*
* Plugin authors export this as the default export from their worker
* entrypoint. The host imports it and calls the lifecycle methods.
*
* @see PLUGIN_SPEC.md §14 — SDK Surface
*/
export interface PaperclipPlugin {
/** The original plugin definition passed to `definePlugin()`. */
readonly definition: PluginDefinition;
}
// ---------------------------------------------------------------------------
// definePlugin — top-level factory
// ---------------------------------------------------------------------------
/**
* Define a Paperclip plugin.
*
* Call this function in your worker entrypoint and export the result as the
* default export. The host will import the module and call lifecycle methods
* on the returned object.
*
* @param definition - Plugin lifecycle handlers
* @returns A sealed `PaperclipPlugin` object for the host to consume
*
* @example
* ```ts
* import { definePlugin } from "@paperclipai/plugin-sdk";
*
* export default definePlugin({
* async setup(ctx) {
* ctx.logger.info("Plugin started");
* ctx.events.on("issue.created", async (event) => {
* // handle event
* });
* },
*
* async onHealth() {
* return { status: "ok" };
* },
* });
* ```
*
* @see PLUGIN_SPEC.md §14.1 — Example SDK Shape
*/
export function definePlugin(definition: PluginDefinition): PaperclipPlugin {
return Object.freeze({ definition });
}