Custom connections are gone: the CLI talks to the LLeMbas server it is logged in to

LLeMbas CLI 2.0 is LLeMbas-server-specific (CLI rule 2). Removed: the
anthropic, gemini, ollama and responses clients, context discovery and
llama-swap unloading (discover.ts), the Connection schema with its dialect,
auth, key_cmd, discover, unload_url and one_model_at_a_time, the model
`fallback` (it only ever moved to another server), effort_map and cache, and
the multi-connection refs. Kept: openai-chat, http, sse, think, effort,
tokens and all of src/lembas.

- connections.yaml holds the login (`type: webui`, as lembas login writes
  it). A custom entry from 1.x is read past with a warning that says how to
  log in, so an old file still starts; with several logins left, the latest
  is used and the others named.
- A model's ref is the id the server serves it under; the older forms
  (`<login>/<id>`, `<login>/<model>`) still resolve and config.yaml is
  rewritten once. Without a login there is no model, and the error (and the
  first-run screen, now the sign-in alone) says `lembas login <address>`.
- A login to another server signs this machine out of the previous one.
- The window is the server's word (no discovery); a capacity group comes
  from the server's `capacity.group` alone; embeddings use the server's
  /v1/embeddings.
- `lembas models` lists the server's models; install.sh's starter files, the
  README and the JSON schemas say the same.

Tests: every fixture is a login now (the fake server lists models, answers
other GETs with 404, and can hand one model to another script, as a server
with several backends); the dialect suites and the cross-connection tests are
gone with their code.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

tests/replay.test.ts

@@ -18,7 +18,7 @@ afterEach(() => fake?.stop())
 
 async function replay(name: string) {
   fake = fakeProvider([{ chunks: [], raw: fixture(name) }])
-  const m: ResolvedModel = { ref: "swap/x", connectionName: "swap", id: "x", spec: {}, connection: { dialect: "openai-chat", base_url: fake.url, models: {} } }
+  const m: ResolvedModel = { ref: "swap/x", connectionName: "swap", id: "x", spec: {}, connection: { base_url: fake.url, models: {} } }
   const events: StreamEvent[] = []
   for await (const e of new OpenAIChatClient(m).stream({ system: "", messages: [{ role: "user", parts: [{ type: "text", text: "x" }] }], tools: [] }))
     events.push(e)
@@ -48,7 +48,7 @@ describe("recorded llama-swap streams", () => {
   test("qwen35: an answer given only inside reasoning is followed by one request for the answer", async () => {
     fake = fakeProvider([{ chunks: [], raw: fixture("qwen35-answer-only-in-reasoning.sse") }, { chunks: [delta({ content: "The divisor was off by one." })] }])
     mkdirSync(paths.config, { recursive: true })
-    writeFileSync(join(paths.config, "connections.yaml"), `connections:\n  f:\n    dialect: openai-chat\n    base_url: ${fake.url}\n    models: { m: {} }\n`, { mode: 0o600 })
+    writeFileSync(join(paths.config, "connections.yaml"), `connections:\n  f:\n    type: webui\n    url: ${fake.url}\n    models: { f/m: {} }\n`, { mode: 0o600 })
     writeFileSync(join(paths.config, "config.yaml"), "model: f/m\n")
     const dir = mkdtempSync(join(tmpdir(), "lembas-replay-"))
     const app = createApp({ cwd: dir, mode: "edit", store: false, asker: { ask: async (): Promise<AskReply> => ({ kind: "deny" }) } })
@@ -95,7 +95,7 @@ describe("effort refused inside a 200 stream (llama-swap loading, then the templ
       connectionName: "swap",
       id: "bonsai",
       spec: { efforts: ["low", "medium", "high"] },
-      connection: { dialect: "openai-chat", base_url: fake.url, models: {} },
+      connection: { base_url: fake.url, models: {} },
     }
     const events: StreamEvent[] = []
     for await (const e of new OpenAIChatClient(m).stream({ system: "", messages: [{ role: "user", parts: [{ type: "text", text: "x" }] }], tools: [], effort: "high" }))
@@ -111,7 +111,7 @@ describe("effort refused inside a 200 stream (llama-swap loading, then the templ
 
   test("an error after output has started is not retried", async () => {
     fake = fakeProvider([{ chunks: [], raw: 'data: {"choices":[{"delta":{"content":"partial"}}]}\n\nerror: {"error":{"code":500,"message":"Unexpected reasoning effort high"}}\n\n' }])
-    const m: ResolvedModel = { ref: "swap/late", connectionName: "swap", id: "x", spec: { efforts: ["high"] }, connection: { dialect: "openai-chat", base_url: fake.url, models: {} } }
+    const m: ResolvedModel = { ref: "swap/late", connectionName: "swap", id: "x", spec: { efforts: ["high"] }, connection: { base_url: fake.url, models: {} } }
     const run = async () => {
       for await (const _ of new OpenAIChatClient(m).stream({ system: "", messages: [], tools: [], effort: "high" })) void _
     }
@@ -123,7 +123,7 @@ describe("effort refused inside a 200 stream (llama-swap loading, then the templ
 describe("llama-swap unloads the model under a running request", () => {
   test("'group: model unloaded' (a string code) is a server error: retried once, and the reply arrives", async () => {
     fake = fakeProvider([{ chunks: [], raw: fixture("llama-swap-model-unloaded.sse") }, { chunks: [delta({ content: "second try" })] }])
-    const m: ResolvedModel = { ref: "swap/x", connectionName: "swap", id: "x", spec: {}, connection: { dialect: "openai-chat", base_url: fake.url, models: {} } }
+    const m: ResolvedModel = { ref: "swap/x", connectionName: "swap", id: "x", spec: {}, connection: { base_url: fake.url, models: {} } }
     const events: StreamEvent[] = []
     for await (const e of new OpenAIChatClient(m).stream({ system: "", messages: [{ role: "user", parts: [{ type: "text", text: "x" }] }], tools: [] })) events.push(e)
     expect(events.some((e) => e.type === "notice" && e.message.includes("group: model unloaded"))).toBe(true)