This article is derived from the plugin's already-indexed fields (wiki / faq / compatibility_json / readme), not freshly generated by AI. Source field is noted at the end of each section.

Quick start

dsh-crew

— source: plugin_wiki.wiki_content

Install & verify

dsh plugin --profile web add @zseven-w/dsh-crew

Run the command above in your DSH Web Profile. Then enable the plugin in the plugin list.

— source: plugins.install

Key points

• DSH (DeepSeek Harness): DeepSeek's open-source agent harness, a code agent in Web UI form, similar to Claude Code but driving DeepSeek models.
• MCP (Model Context Protocol): Anthropic's AI tool integration protocol, enables LLMs to safely call external tools and data sources.
• Cordis bundle: DSH's plugin format; this project can run standalone as an MCP service or install into DSH Web as hub mode.
• worker: the DSH agent doing the work — a full session with its own tools, sandbox and preset, not a bare model call.
• effort: reasoning strength, off = no reasoning, high = high reasoning investment, max = maximum reasoning investment.

— source: plugin_wiki.readme_en (fallback readme_raw)

FAQ

What additional configuration is needed after installation?

If DSH Web is running (hub mode), it already has DeepSeek credentials, and the plugin works directly. Only in pure standalone mode without a DSH instance do you need to write DEEPSEEK_API_KEY to ~/.config/dsh-crew/.env.

DeepSeek is a text-only model, how about images and image generation?

The plugin has two built-in tools: describe_image and generate_image, which use locally logged-in subscription CLIs (Claude, Codex, Grok, Antigravity) or any OpenAI-compatible API/local commands to handle image viewing and generation; pasted images are automatically transcribed to text before being sent to the model.

How to choose between dsh_run_worker and dsh_spawn_worker?

dsh_run_worker is a blocking call that returns after getting the result; dsh_spawn_worker immediately returns a task ID, suitable for parallel fan-out, then use dsh_worker_status / dsh_worker_result to follow the results.

What's the difference between flash and pro tiers?

flash runs faster and cheaper, used for mechanical tasks; pro has stronger reasoning, suitable for complex tasks. You can use tier_policy to direct all dispatches to a single tier, and also allow failed flash to automatically retry with pro once.

How to uninstall?

Click the uninstall button in the DSH Crew section of the DSH settings page, or use the command line node src/install/cli.mjs uninstall; this will clean up Claude Code / Codex registration and HUD segments, while keeping a backup of settings.json.

Which operating system does it need to run on?

Uses only Node built-in modules and subprocesses; can run on any desktop system that can run Node, across macOS / Linux / Windows.

— source: plugin_wiki.faq_json

Compatibility

• DSH: 0.1.0-rc.6+
• Node: 未声明

— source: plugin_wiki.compatibility_json

Pitfalls

Review the upstream repo before installing. This guide is auto-derived from indexed fields and may lag the latest release. If anything contradicts the official docs, treat the upstream source as authoritative.

— source: general rule