2/2 Implement explicit model gateway by monadoid · Pull Request #2440 · browserbase/stagehand

monadoid · 2026-07-26T19:51:09Z

Why

#2438 makes model routing explicit in the protocol. This PR implements that contract in the runtime.

Model IDs from integrated providers are intentionally open: openai/gpt-future-preview is valid direct-provider input without a Stagehand catalog update. The runtime still validates the stable configuration choices it owns: direct credentials, Browserbase gateway selection, and custom endpoint configuration.

What changed

Adds one server-side resolver with three execution targets:

openai/gpt-future-preview
  → direct provider inference

browserbase/openai/gpt-future-preview
  → Browserbase gateway → openai/gpt-future-preview

{ baseURL, modelName }
  → OpenAI-compatible endpoint

Dispatches on the resolved target instead of treating a missing provider API key as Browserbase routing.
Adds the Browserbase gateway client, which:
- derives the regional Stagehand API URL;
- authenticates with the Browserbase API key and session ID;
- forwards the upstream provider-qualified model name unchanged.
Implements the OpenAI-compatible endpoint path.
Updates direct-provider adapters to forward an open provider model ID.
Removes all runtime handling of type: "unlisted".

Test plan

Focused schema, resolver, gateway-dispatch, and runtime-state tests pass.
Gateway behavior is covered with mocked transport and real request construction.
Live Browserbase gateway coverage remains an opt-in integration test because it requires dedicated credentials and a deployed endpoint.

changeset-bot · 2026-07-26T19:51:13Z

⚠️ No Changeset found

Latest commit: 7f253fa

Merging this PR will not cause a version bump for any packages. If these changes should not result in a new version, you're good to go. If these changes should result in a version bump, you need to add a changeset.

Click here to learn what changesets are, and how to add one.

Click here if you're a maintainer who wants to add a changeset to this PR

cubic-dev-ai

12 issues found and verified against the latest diff

Confidence score: 2/5

packages/server/llm/gatewayClient.ts is pointed at a non-deployed .../v1/llm/responses route, so Browserbase model generations will 404 at runtime and effectively take the gateway path down — switch provider configuration to the deployed Model Gateway Responses endpoint before merging.
packages/sdk-python/src/stagehand/_generated/models.py introduced incompatible model config changes (KnownModelConfig.api_key now required and browserbase/auto string construction failing), which can block SDK initialization and make gateway access unusable for Python clients — restore optional api_key handling in validation and route browserbase/ strings through the Browserbase config types.
packages/server/services/llmService.ts always selecting .responses(...) means Chat Completions-compatible custom endpoints cannot generate, creating user-facing regressions for advertised OpenAI-compatible integrations — branch to .chat(...) for those providers.
Coverage gaps across openAiCompatibleClient.ts, stagehandApi.ts, stagehandController.ts, and the act/observe/extract services raise follow-on regression risk around baseURL/apiKey/headers, region routing, and gateway-context forwarding — add focused tests on these new paths to lock behavior before future refactors.

Prompt for AI agents (unresolved issues)


Check if these issues are valid — if so, understand the root cause of each and fix them. If appropriate, use sub-agents to investigate and fix each issue separately.


<file name="packages/server/llm/openAiCompatibleClient.ts">

<violation number="1" location="packages/server/llm/openAiCompatibleClient.ts:6">
P3: Custom endpoint connection settings lack coverage, so regressions that drop or misroute `baseURL`, optional `apiKey`, or custom `headers` will still pass the gateway test. Add focused tests that mock `createOpenAI` and verify these settings for authenticated and header-only endpoints.

(Based on your team's feedback about unit tests for new behavior.) [FEEDBACK_USED]</violation>
</file>

<file name="packages/server/clients/stagehandApi.ts">

<violation number="1" location="packages/server/clients/stagehandApi.ts:14">
P3: Regional endpoint routing has no executable coverage. Add focused cases for each `BrowserbaseRegion` and `undefined`, since a typo or fallback regression sends sessions to a different region and the API rejects the request.</violation>
</file>

<file name="packages/sdk-python/src/stagehand/_generated/models.py">

<violation number="1" location="packages/sdk-python/src/stagehand/_generated/models.py:228">
P1: `Stagehand(model="browserbase/auto")` cannot initialize, leaving the gateway unusable through the SDK's string-model constructor. Route `browserbase/` model strings through `BrowserbaseModelConfig`/`BrowserbaseUnlistedModelConfig` instead of always constructing `KnownModelConfig`.</violation>

<violation number="2" location="packages/sdk-python/src/stagehand/_generated/models.py:772">
P0: `KnownModelConfig.api_key` changed from optional (`Optional[StrictStr] = None`) to required (`StrictStr`), but the SDK's `_model_config()` helper in `client_models.py` explicitly omits `api_key` from the validation dict when no API key is provided. This means any existing code path that creates a model config without an API key — such as calling `Stagehand(model="openai/gpt-5")` without `model_api_key` (line 331 of `stagehand.py` passes `api_key=model_api_key` which defaults to `None`) — will now raise a `ValidationError` at runtime. Recommend keeping `api_key` optional on `KnownModelConfig` or updating `_model_config()` to handle the required field.</violation>
</file>

<file name="packages/server/services/actService.ts">

<violation number="1" location="packages/server/services/actService.ts:243">
P3: Act gateway forwarding has no service-level regression test; add focused coverage that `act()` with a Browserbase model forwards its gateway context (including the two-step/self-heal paths if applicable).

(Based on your team's feedback about unit tests for changed behavior.) [FEEDBACK_USED]</violation>
</file>

<file name="packages/server/controllers/stagehandController.ts">

<violation number="1" location="packages/server/controllers/stagehandController.ts:50">
P3: Gateway routing through the RPC `act`, `observe`, and `extract` paths is untested; current tests only cover `llmService.generate` directly. Add focused controller coverage so a future omitted or misplaced `gateway` argument does not silently break gateway models for these public operations.

(Based on your team's feedback about unit tests for new behavior.) [FEEDBACK_USED]</violation>
</file>

<file name="packages/server/services/extractService.ts">

<violation number="1" location="packages/server/services/extractService.ts:112">
P3: Gateway-backed extraction has no service-level regression coverage, so a future change can silently drop the context and break Browserbase extraction. Add focused tests for forwarding `GatewayContext` and for the missing-context failure path.

(Based on your team's feedback about adding unit tests for changed behavior.) [FEEDBACK_USED]</violation>
</file>

<file name="packages/server/services/observeService.ts">

<violation number="1" location="packages/server/services/observeService.ts:89">
P3: Gateway-backed `observe()` routing has no service-level regression test. Add a focused observation test with a Browserbase model and gateway context so this new forwarding path remains covered.

(Based on your team's feedback about adding unit tests for new behavior.) [FEEDBACK_USED]</violation>
</file>

<file name="packages/server/llm/aiSdkClient.ts">

<violation number="1" location="packages/server/llm/aiSdkClient.ts:125">
P3: Unlisted direct models now reach provider factories without catalog validation, but no AI SDK-client test covers that new path. Add focused unlisted-ID coverage (including a slash-containing ID) so forwarding behavior cannot regress.

(Based on your team's feedback about unit tests for new behavior.) [FEEDBACK_USED]</violation>
</file>

<file name="packages/server/services/llmService.ts">

<violation number="1" location="packages/server/services/llmService.ts:33">
P2: Custom endpoints that expose the common Chat Completions-compatible API cannot generate here: this route always selects the helper's `.responses(...)` model. Use `.chat(...)` for the advertised OpenAI-compatible contract, or explicitly restrict and document custom endpoints as Responses-compatible.</violation>
</file>

<file name="packages/server/llm/gatewayClient.ts">

<violation number="1" location="packages/server/llm/gatewayClient.ts:25">
P3: Gateway transport wiring has no focused test, so an invalid endpoint or missing cloud headers reaches production undetected. Add tests that mock the provider request and assert the resolved URL plus both Browserbase headers.

(Based on your team's feedback about unit tests for new behavior.) [FEEDBACK_USED]</violation>

<violation number="2" location="packages/server/llm/gatewayClient.ts:31">
P0: All Browserbase model generations will 404 because this produces `<regional API>/v1/llm/responses`, which is not a deployed route. Configure the provider with the deployed Model Gateway Responses endpoint before dispatching requests.</violation>
</file>

_{Reply with feedback, questions, or to request a fix.

Re-trigger cubic}

cubic-dev-ai · 2026-07-26T20:02:43Z

+): LanguageModel {
+  return createOpenAI({
+    apiKey: gateway.apiKey,
+    baseURL: `${gateway.apiUrl}/llm`,


P0: All Browserbase model generations will 404 because this produces <regional API>/v1/llm/responses, which is not a deployed route. Configure the provider with the deployed Model Gateway Responses endpoint before dispatching requests.

Prompt for AI agents

Check if this issue is valid — if so, understand the root cause and fix it. At packages/server/llm/gatewayClient.ts, line 31: <comment>All Browserbase model generations will 404 because this produces `<regional API>/v1/llm/responses`, which is not a deployed route. Configure the provider with the deployed Model Gateway Responses endpoint before dispatching requests.</comment> <file context> @@ -0,0 +1,37 @@ +): LanguageModel { + return createOpenAI({ + apiKey: gateway.apiKey, + baseURL: `${gateway.apiUrl}/llm`, + headers: { + "x-bb-api-key": gateway.apiKey, </file context>

cubic-dev-ai · 2026-07-26T20:02:43Z

    api_key: Annotated[
-        Optional[StrictStr], Field(examples=["sk-some-openai-api-key"], min_length=1)
-    ] = None
+        StrictStr, Field(examples=["sk-some-openai-api-key"], min_length=1)


P0: KnownModelConfig.api_key changed from optional (Optional[StrictStr] = None) to required (StrictStr), but the SDK's _model_config() helper in client_models.py explicitly omits api_key from the validation dict when no API key is provided. This means any existing code path that creates a model config without an API key — such as calling Stagehand(model="openai/gpt-5") without model_api_key (line 331 of stagehand.py passes api_key=model_api_key which defaults to None) — will now raise a ValidationError at runtime. Recommend keeping api_key optional on KnownModelConfig or updating _model_config() to handle the required field.

Prompt for AI agents

Check if this issue is valid — if so, understand the root cause and fix it. At packages/sdk-python/src/stagehand/_generated/models.py, line 772: <comment>`KnownModelConfig.api_key` changed from optional (`Optional[StrictStr] = None`) to required (`StrictStr`), but the SDK's `_model_config()` helper in `client_models.py` explicitly omits `api_key` from the validation dict when no API key is provided. This means any existing code path that creates a model config without an API key — such as calling `Stagehand(model="openai/gpt-5")` without `model_api_key` (line 331 of `stagehand.py` passes `api_key=model_api_key` which defaults to `None`) — will now raise a `ValidationError` at runtime. Recommend keeping `api_key` optional on `KnownModelConfig` or updating `_model_config()` to handle the required field.</comment> <file context> @@ -725,8 +769,8 @@ class KnownModelConfig(WireModel): api_key: Annotated[ - Optional[StrictStr], Field(examples=["sk-some-openai-api-key"], min_length=1) - ] = None + StrictStr, Field(examples=["sk-some-openai-api-key"], min_length=1) + ] """ </file context>

cubic-dev-ai · 2026-07-26T20:02:43Z

    min_width: Optional[StrictFloat] = None


+class BrowserbaseModelConfig(WireModel):


P1: Stagehand(model="browserbase/auto") cannot initialize, leaving the gateway unusable through the SDK's string-model constructor. Route browserbase/ model strings through BrowserbaseModelConfig/BrowserbaseUnlistedModelConfig instead of always constructing KnownModelConfig.

Prompt for AI agents

Check if this issue is valid — if so, understand the root cause and fix it. At packages/sdk-python/src/stagehand/_generated/models.py, line 228: <comment>`Stagehand(model="browserbase/auto")` cannot initialize, leaving the gateway unusable through the SDK's string-model constructor. Route `browserbase/` model strings through `BrowserbaseModelConfig`/`BrowserbaseUnlistedModelConfig` instead of always constructing `KnownModelConfig`.</comment> <file context> @@ -225,6 +225,31 @@ class BrowserbaseFingerprintScreen(WireModel): min_width: Optional[StrictFloat] = None +class BrowserbaseModelConfig(WireModel): + model_config = ConfigDict( + extra="forbid", </file context>

cubic-dev-ai · 2026-07-26T20:02:43Z

+    case "direct":
+      return await generateWithAiSdk(createAiSdkLanguageModel(target), params);
+    case "openai-compatible":
+      return await generateWithAiSdk(createOpenAICompatibleLanguageModel(target), params);


P2: Custom endpoints that expose the common Chat Completions-compatible API cannot generate here: this route always selects the helper's .responses(...) model. Use .chat(...) for the advertised OpenAI-compatible contract, or explicitly restrict and document custom endpoints as Responses-compatible.

Prompt for AI agents

Check if this issue is valid — if so, understand the root cause and fix it. At packages/server/services/llmService.ts, line 33: <comment>Custom endpoints that expose the common Chat Completions-compatible API cannot generate here: this route always selects the helper's `.responses(...)` model. Use `.chat(...)` for the advertised OpenAI-compatible contract, or explicitly restrict and document custom endpoints as Responses-compatible.</comment> <file context> @@ -7,27 +7,34 @@ import type { + case "direct": + return await generateWithAiSdk(createAiSdkLanguageModel(target), params); + case "openai-compatible": + return await generateWithAiSdk(createOpenAICompatibleLanguageModel(target), params); + case "browserbase": + if (!gateway) { </file context>

cubic-dev-ai · 2026-07-26T20:02:43Z

+import type { OpenAICompatibleModelTarget } from "./modelTarget.js";
+
+/** Creates an OpenAI Responses-compatible model for a caller-controlled endpoint. */
+export function createOpenAICompatibleLanguageModel(


P3: Custom endpoint connection settings lack coverage, so regressions that drop or misroute baseURL, optional apiKey, or custom headers will still pass the gateway test. Add focused tests that mock createOpenAI and verify these settings for authenticated and header-only endpoints.

(Based on your team's feedback about unit tests for new behavior.)

View Feedback

Prompt for AI agents

Check if this issue is valid — if so, understand the root cause and fix it. At packages/server/llm/openAiCompatibleClient.ts, line 6: <comment>Custom endpoint connection settings lack coverage, so regressions that drop or misroute `baseURL`, optional `apiKey`, or custom `headers` will still pass the gateway test. Add focused tests that mock `createOpenAI` and verify these settings for authenticated and header-only endpoints. (Based on your team's feedback about unit tests for new behavior.) </comment> <file context> @@ -0,0 +1,14 @@ +import type { OpenAICompatibleModelTarget } from "./modelTarget.js"; + +/** Creates an OpenAI Responses-compatible model for a caller-controlled endpoint. */ +export function createOpenAICompatibleLanguageModel( + target: OpenAICompatibleModelTarget, +): LanguageModel { </file context>

cubic-dev-ai · 2026-07-26T20:02:43Z

      selfHeal: state.initParams.selfHeal,
      domSettleTimeoutMs: state.initParams.domSettleTimeoutMs,
      cache: cacheService.buildCacheContext(state.initParams),
+      gateway: buildGatewayContext(state.initParams),


P3: Gateway routing through the RPC act, observe, and extract paths is untested; current tests only cover llmService.generate directly. Add focused controller coverage so a future omitted or misplaced gateway argument does not silently break gateway models for these public operations.

(Based on your team's feedback about unit tests for new behavior.)

View Feedback

Prompt for AI agents

Check if this issue is valid — if so, understand the root cause and fix it. At packages/server/controllers/stagehandController.ts, line 50: <comment>Gateway routing through the RPC `act`, `observe`, and `extract` paths is untested; current tests only cover `llmService.generate` directly. Add focused controller coverage so a future omitted or misplaced `gateway` argument does not silently break gateway models for these public operations. (Based on your team's feedback about unit tests for new behavior.) </comment> <file context> @@ -46,6 +47,7 @@ export function createStagehandController(runtime: StagehandRuntime) { selfHeal: state.initParams.selfHeal, domSettleTimeoutMs: state.initParams.domSettleTimeoutMs, cache: cacheService.buildCacheContext(state.initParams), + gateway: buildGatewayContext(state.initParams), }); } </file context>

cubic-dev-ai · 2026-07-26T20:02:43Z

        domElements: combinedTree,
        schema: transformedSchema as z.ZodObject,
-        generate: (input) => llmService.generate(model, input, clientLLMGenerate),
+        generate: (input) => llmService.generate(model, input, clientLLMGenerate, gateway),


P3: Gateway-backed extraction has no service-level regression coverage, so a future change can silently drop the context and break Browserbase extraction. Add focused tests for forwarding GatewayContext and for the missing-context failure path.

(Based on your team's feedback about adding unit tests for changed behavior.)

View Feedback

Prompt for AI agents

Check if this issue is valid — if so, understand the root cause and fix it. At packages/server/services/extractService.ts, line 112: <comment>Gateway-backed extraction has no service-level regression coverage, so a future change can silently drop the context and break Browserbase extraction. Add focused tests for forwarding `GatewayContext` and for the missing-context failure path. (Based on your team's feedback about adding unit tests for changed behavior.) </comment> <file context> @@ -106,7 +109,7 @@ export async function extract({ domElements: combinedTree, schema: transformedSchema as z.ZodObject, - generate: (input) => llmService.generate(model, input, clientLLMGenerate), + generate: (input) => llmService.generate(model, input, clientLLMGenerate, gateway), userProvidedInstructions: systemPrompt, }); </file context>

cubic-dev-ai · 2026-07-26T20:02:43Z

      instruction: effectiveInstruction,
      domElements: combinedTree,
-      generate: (input) => llmService.generate(model, input, clientLLMGenerate),
+      generate: (input) => llmService.generate(model, input, clientLLMGenerate, gateway),


P3: Gateway-backed observe() routing has no service-level regression test. Add a focused observation test with a Browserbase model and gateway context so this new forwarding path remains covered.

(Based on your team's feedback about adding unit tests for new behavior.)

View Feedback

Prompt for AI agents

Check if this issue is valid — if so, understand the root cause and fix it. At packages/server/services/observeService.ts, line 89: <comment>Gateway-backed `observe()` routing has no service-level regression test. Add a focused observation test with a Browserbase model and gateway context so this new forwarding path remains covered. (Based on your team's feedback about adding unit tests for new behavior.) </comment> <file context> @@ -83,7 +86,7 @@ export async function observe({ instruction: effectiveInstruction, domElements: combinedTree, - generate: (input) => llmService.generate(model, input, clientLLMGenerate), + generate: (input) => llmService.generate(model, input, clientLLMGenerate, gateway), userProvidedInstructions: systemPrompt, supportedActions: Object.values(SupportedUnderstudyAction), </file context>

cubic-dev-ai · 2026-07-26T20:02:43Z


 /** Creates a direct AI SDK model from a validated Stagehand model configuration. */
-export function createAiSdkLanguageModel(config: KnownModelConfig): LanguageModel {
+export function createAiSdkLanguageModel(config: DirectModelConnection): LanguageModel {


P3: Unlisted direct models now reach provider factories without catalog validation, but no AI SDK-client test covers that new path. Add focused unlisted-ID coverage (including a slash-containing ID) so forwarding behavior cannot regress.

(Based on your team's feedback about unit tests for new behavior.)

View Feedback

Prompt for AI agents

Check if this issue is valid — if so, understand the root cause and fix it. At packages/server/llm/aiSdkClient.ts, line 125: <comment>Unlisted direct models now reach provider factories without catalog validation, but no AI SDK-client test covers that new path. Add focused unlisted-ID coverage (including a slash-containing ID) so forwarding behavior cannot regress. (Based on your team's feedback about unit tests for new behavior.) </comment> <file context> @@ -128,7 +122,7 @@ const AiSdkGenerationSchema = z /** Creates a direct AI SDK model from a validated Stagehand model configuration. */ -export function createAiSdkLanguageModel(config: KnownModelConfig): LanguageModel { +export function createAiSdkLanguageModel(config: DirectModelConnection): LanguageModel { const separator = config.modelName.indexOf("/"); const provider = ModelProviderSchema.parse(config.modelName.slice(0, separator)); </file context>

cubic-dev-ai · 2026-07-26T20:02:43Z

+}
+
+/** Creates an AI SDK Responses model backed by Browserbase Model Gateway. */
+export function createGatewayLanguageModel(


P3: Gateway transport wiring has no focused test, so an invalid endpoint or missing cloud headers reaches production undetected. Add tests that mock the provider request and assert the resolved URL plus both Browserbase headers.

(Based on your team's feedback about unit tests for new behavior.)

View Feedback

Prompt for AI agents

Check if this issue is valid — if so, understand the root cause and fix it. At packages/server/llm/gatewayClient.ts, line 25: <comment>Gateway transport wiring has no focused test, so an invalid endpoint or missing cloud headers reaches production undetected. Add tests that mock the provider request and assert the resolved URL plus both Browserbase headers. (Based on your team's feedback about unit tests for new behavior.) </comment> <file context> @@ -0,0 +1,37 @@ +} + +/** Creates an AI SDK Responses model backed by Browserbase Model Gateway. */ +export function createGatewayLanguageModel( + target: BrowserbaseModelTarget, + gateway: GatewayContext, </file context>

monadoid changed the title ~~Implement model gateway~~ Implement explicit model gateway Jul 26, 2026

monadoid force-pushed the implement-model-gateway branch from d2b2604 to 7f253fa Compare July 26, 2026 20:01

cubic-dev-ai Bot reviewed Jul 26, 2026

View reviewed changes

monadoid added the v4-pre-release label Jul 27, 2026

monadoid mentioned this pull request Jul 27, 2026

just add-model workflow #2442

Closed

monadoid added 3 commits July 27, 2026 22:36

Add model routing protocol

a749a31

Implement model gateway

e916922

Accept open provider model names at runtime

63fbb89

monadoid force-pushed the implement-model-gateway branch from 7f253fa to 63fbb89 Compare July 27, 2026 20:38

monadoid force-pushed the add-explicit-model-gateway-protocol branch from beb241b to 5fba8ca Compare July 27, 2026 20:38

monadoid mentioned this pull request Jul 28, 2026

1/2 Add gateway concept to protocol #2438

Open

monadoid changed the title ~~Implement explicit model gateway~~ 2/2 Implement explicit model gateway Jul 28, 2026

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

2/2 Implement explicit model gateway - #2440

2/2 Implement explicit model gateway#2440
monadoid wants to merge 3 commits into
add-explicit-model-gateway-protocolfrom
implement-model-gateway

monadoid commented Jul 26, 2026 •

edited

Loading

Uh oh!

changeset-bot Bot commented Jul 26, 2026 •

edited

Loading

Uh oh!

cubic-dev-ai Bot left a comment

Uh oh!

cubic-dev-ai Bot Jul 26, 2026

Uh oh!

cubic-dev-ai Bot Jul 26, 2026

Uh oh!

cubic-dev-ai Bot Jul 26, 2026

Uh oh!

cubic-dev-ai Bot Jul 26, 2026

Uh oh!

cubic-dev-ai Bot Jul 26, 2026

Uh oh!

cubic-dev-ai Bot Jul 26, 2026

Uh oh!

cubic-dev-ai Bot Jul 26, 2026

Uh oh!

cubic-dev-ai Bot Jul 26, 2026

Uh oh!

cubic-dev-ai Bot Jul 26, 2026

Uh oh!

cubic-dev-ai Bot Jul 26, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

		min_width: Optional[StrictFloat] = None


		class BrowserbaseModelConfig(WireModel):

Uh oh!

Conversation

monadoid commented Jul 26, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Why

What changed

Test plan

Uh oh!

changeset-bot Bot commented Jul 26, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

⚠️ No Changeset found

Uh oh!

cubic-dev-ai Bot left a comment

Choose a reason for hiding this comment

Uh oh!

cubic-dev-ai Bot Jul 26, 2026

Choose a reason for hiding this comment

Uh oh!

cubic-dev-ai Bot Jul 26, 2026

Choose a reason for hiding this comment

Uh oh!

cubic-dev-ai Bot Jul 26, 2026

Choose a reason for hiding this comment

Uh oh!

cubic-dev-ai Bot Jul 26, 2026

Choose a reason for hiding this comment

Uh oh!

cubic-dev-ai Bot Jul 26, 2026

Choose a reason for hiding this comment

Uh oh!

cubic-dev-ai Bot Jul 26, 2026

Choose a reason for hiding this comment

Uh oh!

cubic-dev-ai Bot Jul 26, 2026

Choose a reason for hiding this comment

Uh oh!

cubic-dev-ai Bot Jul 26, 2026

Choose a reason for hiding this comment

Uh oh!

cubic-dev-ai Bot Jul 26, 2026

Choose a reason for hiding this comment

Uh oh!

cubic-dev-ai Bot Jul 26, 2026

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

monadoid commented Jul 26, 2026 •

edited

Loading

changeset-bot Bot commented Jul 26, 2026 •

edited

Loading