feat(cli): add orchestrator provider to init wizard

CHANGELOG.md

@@ -6,6 +6,9 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/).
 
 ## [Unreleased]
 
+### Added
+- Orchestrator provider option in `zeph init` wizard for multi-model routing setup (#597)
+
 ### Fixed
 - Restore `--vault`, `--vault-key`, `--vault-path` CLI flags lost during clap migration (#587)
 

README.md

@@ -69,6 +69,9 @@ zeph
 ZEPH_LLM_PROVIDER=claude ZEPH_CLAUDE_API_KEY=sk-ant-... zeph
 ZEPH_LLM_PROVIDER=openai ZEPH_OPENAI_API_KEY=sk-... zeph
 
+# Multi-model routing — primary Claude, fallback Ollama
+ZEPH_LLM_PROVIDER=orchestrator zeph  # configure via `zeph init`
+
 # Any OpenAI-compatible API (Together AI, Groq, Fireworks, etc.)
 ZEPH_LLM_PROVIDER=compatible ZEPH_COMPATIBLE_BASE_URL=https://api.together.xyz/v1 \
   ZEPH_COMPATIBLE_API_KEY=... zeph

docs/src/changelog.md

@@ -8,6 +8,9 @@ See the full [CHANGELOG.md](https://github.com/bug-ops/zeph/blob/main/CHANGELOG.
 
 ## [Unreleased]
 
+### Added
+- Orchestrator provider option in `zeph init` wizard for multi-model routing setup (#597)
+
 ## [0.11.0] - 2026-02-19
 
 ### Added

docs/src/getting-started/configuration.md

@@ -4,7 +4,7 @@
 
 Run `zeph init` to generate a `config.toml` interactively. The wizard walks through five steps:
 
-1. **LLM Provider** -- select Ollama (local), Claude, OpenAI, or a compatible endpoint. Provide the base URL, model name, and API key as needed. Choose an embedding model (default: `qwen3-embedding`).
+1. **LLM Provider** -- select Ollama (local), Claude, OpenAI, Orchestrator (multi-model routing), or a compatible endpoint. Orchestrator prompts for a primary and fallback provider, enabling automatic failover. Provide the base URL, model name, and API key as needed. Choose an embedding model (default: `qwen3-embedding`).
 2. **Memory** -- set the SQLite database path and optionally enable semantic memory with Qdrant.
 3. **Channel** -- pick CLI (default), Telegram, Discord, or Slack. Provide tokens and credentials for the selected channel.
 4. **Secrets backend** -- choose `env` (environment variables) or `age` (encrypted file via `~/.zeph/vault.age`).

docs/src/guide/orchestrator.md

@@ -51,6 +51,16 @@ Routes define provider preference order. If the first provider fails, the next o
 coding = ["local", "cloud"]  # try local first, fallback to cloud
 ```
 
+## Interactive Setup
+
+Run `zeph init` and select **Orchestrator** as the LLM provider. The wizard prompts for:
+
+1. **Primary provider** — select from Ollama, Claude, OpenAI, or Compatible. Provide the model name, base URL, and API key as needed.
+2. **Fallback provider** — same selection. The fallback activates when the primary fails.
+3. **Embedding model** — used for skill matching and semantic memory.
+
+The wizard generates a complete `[llm.orchestrator]` section with provider map, `chat` route (primary + fallback), and `embed` route.
+
 ## Hybrid Setup Example
 
 Embeddings via free local Ollama, chat via paid Claude API: