@cyansalt/dsh-ark-profile

Ark models, web search, and browser settings for DeepSeek Harness.

Installation

Install and enable the bundle in the DSH Web profile:

dsh plugin --profile web add @cyansalt/dsh-ark-profile

Usage

Select @cyansalt/dsh-ark-profile as a DSH profile. The bundle configures the Ark provider and mounts the packages needed for web search and browser-based settings:

Entry	Package	Purpose
llm-pi-ai	@deepseek-ai/dsh-llm-pi-ai	Provides the ark model route
web-search-ark	@cyansalt/dsh-web-search-ark	Runs Ark web searches
ui-settings-ark	@cyansalt/dsh-client-ui-settings-ark	Adds Ark settings to the profile page
web	@deepseek-ai/dsh-web	Selects web-search-ark as the search provider

Configure an actual Ark endpoint or model ID before making model requests. The bundled unknown model is only a placeholder that makes Ark available on model configuration surfaces.

Defaults

The profile adds llm-pi-ai.providers.ark with these values:

Field	Default
displayName	Ark
api	openai-responses
apiKeyEnv	ARK_API_KEY
baseURL	https://ark.cn-beijing.volces.com/api/v3
models	[{ id: "unknown", name: "Unknown" }]

The web-search and settings entries both use the ark provider profile, so model requests and web-search requests share its endpoint and credential reference. The settings card can edit the endpoint, the referenced API key, and the maximum number of keywords per search. It does not create or replace the provider's model list.

This package only composes the bundle. The runtime implementations remain in the packages listed above.