dsh-plugin-tokenmarket

Wiring a new model provider into DeepSeek Harness usually means configuring a base URL, protocol, credentials, and writing adapter glue code for each one. dsh-plugin-tokenmarket collapses that into a single bundle install: it registers Token Market as an OpenAI-compatible tokenmarket provider, mounts the official pi-ai adapter, and lets multiple model families share one API key. The starter catalog ships with DeepSeek V4 Pro, Kimi K3, and GLM-5.3, kept intentionally small so every bundled model can be verified and kept current.

The day-to-day experience centers on practical inference workflows. Streaming responses keep long-form generation responsive; tool calling works out of the box through the pi-ai adapter without custom glue; and the DeepSeek-compatible thinking format surfaces reasoning steps in a structured way that is easy to debug and display. Live model discovery via GET /v1/models means catalog changes do not require config edits, and credentials stay outside the repository through the TOKENMARKET_API_KEY environment variable.

If you are on the DeepSeek Harness developer preview (0.1.5-rc.1 compatible) and need to compare, fall back, or A/B test across multiple model families within a single profile, this bundle lets you do it with one console and one key, freeing your attention for prompt engineering and application logic.