dsh-llm-ollama

Ollama Cloud integration plugin for DeepSeek Harness: a native chat provider through the shared pi-ai-backed adapter, Ollama-native model discovery, Web Search/Fetch providers, and a Web configuration card under Settings → LLM Providers. The LLM Providers page and shared order store are owned by dsh-llm-providers-ui.

• Native Ollama Cloud provider for DeepSeek Harness; chat uses Ollama's OpenAI-compatible Chat Completions endpoint through the shared pi-ai-backed adapter.
• Contributes an Ollama Cloud card under Settings → LLM Providers via the ./client export; the LLM Providers page is owned by dsh-llm-providers-ui.
• Model discovery uses Ollama-native GET /api/tags and POST /api/show; picker enriches up to six models concurrently and can preserve current-only models.
• Registers Web Search (POST /api/web_search) and Web Fetch (POST /api/web_fetch) as ctx.web providers under ollama-cloud.
• Cloud usage section mirrors ollama.com/settings, rendering each window the endpoint reports as consumed-percentage meters with per-model request counts.
• Configurable apiKeyEnv, baseURL, defaultContextWindow, streamIdleTimeoutMs, webRequestTimeoutMs, retryPolicy and per-model entries.
• Retries eligible model-request failures up to eight times by default; authentication, invalid-request and unsupported-content failures are non-retryable.
• Package root exposes the Cordis plugin contract and OllamaAdapter; release artifact contains built Host/Client files and targets listed Harness releases.