dsh-auxiliary

Configuration

All fields are optional; defaults are shown.

- name: dsh-auxiliary
  config:
    vision:
      maxTokens: 2048                      # inspect_image output cap (provider/model written by the settings page)
    tool:
      enabled: true                        # register the inspect_image tool
      maxImageBytes: 10485760              # per-file size cap
      timeoutMs: 120000                    # cooperative tool-call budget
    compact:
      enabled: false                       # reroute compaction summaries to an auxiliary model
      provider: ""                         # e.g. deepseek-official (a registered provider route id)
      model: ""                            # e.g. deepseek-chat (a model id on that provider)
    engine:
      enabled: false                       # optional compression engine (mutually exclusive with dsh-compaction-basic)
      thresholdRatio: 0.8
      retainRatio: 0.16
      maxTokens: 8192
      compactionRetries: 1
      maxOverflowRetries: 1
      auto: true
      compressPrompt: "..."                # custom compression instruction

Settings page: Auxiliary Models

The plugin ships a web settings section (Settings → Auxiliary Models). Configure a provider and its models in the Models page first, then pick them here: the page lists only providers that are currently enabled and advertise at least one model in the catalog, and the model list shows only that provider's catalog models. Saving writes vision.provider and vision.model — the selection only affects the vision model inspect_image uses; compaction summaries keep their own independent compact.provider / compact.model pair.

Notes

• compact.enabled reroutes only purpose: 'compaction' calls. Point provider/model at any registered route — e.g. deepseek-official with a cheaper model (placeholders; replace with your actual provider route and model id).
• engine.enabled: true replaces the stock compaction backend; do not load @deepseek-ai/dsh-compaction-basic at the same time. The plugin detects the conflict and skips the engine with a warning.
• Vision tool arguments: path (absolute or workspace-relative) and optional question. Supported formats: PNG, JPEG, WebP, GIF.