FitBBC/dsh-plugin-tokenmarket

Token Market for DeepSeek Harness

An installable DeepSeek Harness Profile Bundle that adds Token Market as an OpenAI-compatible model provider.

中文说明

Features

• One Token Market API key for multiple model families.
• Streaming responses and tool calling through the Harness pi-ai adapter.
• Reasoning output through the DeepSeek-compatible thinking format.
• Live model discovery through GET /v1/models.
• Credentials stay outside the repository and configuration file.

The starter catalog includes DeepSeek V4 Pro, Kimi K3, and GLM-5.3. The catalog is intentionally small so every bundled model can be tested and kept current. These models were confirmed available through Token Market on September 10, 2026.

Requirements

• DeepSeek Harness developer preview compatible with 0.1.5-rc.1.
• Node.js ^22.19.0 or >=24.0.0.
• A Token Market API key from the Token Market Console.

Install

Install the bundle into a Harness profile:

dsh plugin --profile web add github:FitBBC/dsh-plugin-tokenmarket

Verify that the bundle is present in the composed profile:

dsh --profile web --dump-config

Set the credential in your environment or a Harness-supported .env file:

TOKENMARKET_API_KEY=your_api_key

Restart the profile after adding the bundle:

dsh --profile web

Token Market will appear as the tokenmarket provider.

Verify the live catalog

TOKENMARKET_API_KEY=your_api_key npm run check:models

This command calls the authenticated Token Market model endpoint and fails if one of the bundled starter models is no longer available.

How it works

This package is a configuration-only Harness Profile Bundle. Its cordis.patch.yml mounts the official @deepseek-ai/dsh-llm-pi-ai adapter and registers a tokenmarket route with:

• API base URL: https://api.tokensmarket.ai/v1
• Protocol: openai-completions
• Credential reference: TOKENMARKET_API_KEY

See the Token Market documentation for API details and the model catalog for current availability and pricing.

Development

npm test
npm pack --dry-run

License

MIT