aigetwey 1.4.2 → 1.4.4

package/dashboard/src/lib/capabilities.ts

@@ -1,318 +0,0 @@
-// Model capabilities — what each model can read/do beyond plain text.
-//
-// Fallback order (first match wins), result merged over DEFAULT_CAPABILITIES:
-//   1. PROVIDER_CAPABILITIES[provider][model]  — provider-specific override
-//   2. MODEL_CAPABILITIES[model]               — canonical exact id (handles exceptions)
-//   3. PATTERN_CAPABILITIES                     — glob match, ordered specific -> generic
-//   4. DEFAULT_CAPABILITIES                     — safe floor (always returned)
-//
-// ── HOW TO ADD / UPDATE A MODEL ──────────────────────────────────────
-// Authoritative data source: https://models.dev/api.json (145 providers, 4000+
-// models, MIT). Each model exposes the exact fields we map below:
-//   modalities.input  ["text","image","pdf","audio","video"] -> vision / pdf / audioInput / videoInput
-//   modalities.output ["text","image","audio"]               -> imageOutput / audioOutput
-//   reasoning   -> reasoning      tool_call    -> tools
-//   limit.context -> contextWindow   limit.output -> maxOutput
-// Look up the model id, then:
-//   • If a PATTERN below already covers it correctly -> nothing to do.
-//   • If it is an exception (pattern would mis-match) -> add an exact entry to
-//     MODEL_CAPABILITIES (only the fields that differ from DEFAULT).
-//   • If a whole new family -> add an ordered PATTERN (specific before generic).
-// NOTE: models.dev has NO "search" flag (web search is a runtime tool, not a
-// model spec); set `search` from vendor docs (Claude 4.x+, GPT-5.x/4o, Gemini
-// 2.0+, Grok, Perplexity). Verify with: curl -s https://models.dev/api.json
-
-
-/**
- * Safe floor — every resolved result is merged over this so consumers
- * never need null-checks. Most modern LLMs meet these limits.
- */
-export interface Caps {
-  vision: boolean;
-  pdf: boolean;
-  audioInput: boolean;
-  videoInput: boolean;
-  imageOutput: boolean;
-  audioOutput: boolean;
-  search: boolean;
-  tools: boolean;
-  reasoning: boolean;
-  thinkingFormat: string | null;
-  thinkingCanDisable: boolean;
-  thinkingRange: { min: number; max: number } | null;
-  contextWindow: number;
-  maxOutput: number;
-}
-
-/**
- * Glob (* = wildcard) match, anchored + case-insensitive. aigetwey's
- * pricing.matchPattern so capabilities resolve identically.
- */
-export function matchPattern(pattern: string, model: string): boolean {
-  const regex = new RegExp(
-    "^" + pattern.split("*").map((s) => s.replace(/[.*+?^${}()|[\]\\]/g, "\\$&")).join(".*") + "$",
-    "i",
-  );
-  return regex.test(model);
-}
-
-export const DEFAULT_CAPABILITIES: Caps = {
-  // input modalities
-  vision: false,        // read images
-  pdf: false,           // read PDF / documents
-  audioInput: false,    // read audio
-  videoInput: false,    // read video
-  // output modalities
-  imageOutput: false,   // generate images
-  audioOutput: false,   // generate audio
-  // features
-  search: false,        // built-in web search tool / grounding
-  tools: true,          // function / tool calling
-  reasoning: false,     // thinking / reasoning
-  // thinking wire format (only meaningful when reasoning:true). null → derive from transport.format.
-  // enum: openai|claude-adaptive|claude-budget|gemini-level|gemini-budget|zai|qwen|deepseek|kimi|minimax|hunyuan|step
-  thinkingFormat: null,
-  thinkingCanDisable: true,  // false → model cannot turn thinking off (clamp to min instead of disable)
-  thinkingRange: null,       // { min, max } for budget formats; null = no clamp
-  // limits (tokens)
-  contextWindow: 200000,
-  maxOutput: 64000,
-};
-
-// User-added model metadata can carry dashboard service kinds instead of the
-// runtime capability names used here. Map those typed model kinds into input /
-// output capabilities so custom vision models are not treated as text-only.
-const SERVICE_KIND_CAPABILITIES: Record<string, Partial<Caps>> = {
-  imageToText: { vision: true },
-  image: { imageOutput: true },
-  stt: { audioInput: true },
-  tts: { audioOutput: true },
-  embedding: { tools: false },
-};
-
-export function capabilitiesFromServiceKind(kind: string): Partial<Caps> | null {
-  return SERVICE_KIND_CAPABILITIES[kind] || null;
-}
-
-/**
- * Canonical exact-id overrides — used for exceptions that patterns would
- * otherwise mis-match. Only declare deltas vs DEFAULT.
- */
-export const MODEL_CAPABILITIES: Record<string, Partial<Caps>> = {
-  // Claude 4.6/4.7 have 1M context + adaptive thinking (override generic claude pattern)
-  "claude-opus-4.6":   { vision: true, reasoning: true, search: true, thinkingFormat: "claude-adaptive", contextWindow: 1000000, maxOutput: 128000 },
-  "claude-opus-4.7":   { vision: true, reasoning: true, search: true, thinkingFormat: "claude-adaptive", contextWindow: 1000000, maxOutput: 128000 },
-  "claude-opus-4-6":   { vision: true, reasoning: true, search: true, thinkingFormat: "claude-adaptive", contextWindow: 1000000, maxOutput: 128000 },
-  "claude-sonnet-4.6": { vision: true, reasoning: true, search: true, thinkingFormat: "claude-adaptive", contextWindow: 1000000, maxOutput: 64000 },
-  "claude-sonnet-4-6": { vision: true, reasoning: true, search: true, thinkingFormat: "claude-adaptive", contextWindow: 1000000, maxOutput: 64000 },
-
-  // Gemini image-gen / OpenAI image / xai image variants
-  "gpt-image-1":       { imageOutput: true, tools: false },
-
-  // GLM vision variant (text GLM has no vision)
-  "glm-4.6v":          { vision: true, reasoning: true, thinkingFormat: "zai", contextWindow: 128000 },
-
-  // Qwen plain coder/text (no vision) — registry "vision-model" / "coder-model" aliases
-  "vision-model":      { vision: true, reasoning: true, thinkingFormat: "qwen", contextWindow: 1000000 },
-  "coder-model":       { reasoning: true, thinkingFormat: "qwen", contextWindow: 1000000 },
-};
-
-/**
- * Provider-specific capability overrides. Keyed by provider alias/id.
- */
-export const PROVIDER_CAPABILITIES: Record<string, Record<string, Partial<Caps>>> = {
-  // CodeBuddy.cn — authoritative per-model metadata from the gateway's model
-  // config (contextWindow=maxInputTokens, maxOutput=maxOutputTokens, vision=
-  // supportsImages). Every model reasons via OpenAI-style reasoning_effort
-  // (see registry thinkingFormat). `onlyReasoning` models can't turn thinking
-  // off → thinkingCanDisable:false (clamped to minimal instead of disabled).
-  "codebuddy-cn": {
-    "glm-5.2":            { reasoning: true, thinkingFormat: "openai", thinkingCanDisable: false, contextWindow: 1000000, maxOutput: 48000 },
-    "glm-5.1":            { reasoning: true, thinkingFormat: "openai", thinkingCanDisable: false, contextWindow: 200000, maxOutput: 48000 },
-    "glm-5.0":            { reasoning: true, thinkingFormat: "openai", contextWindow: 200000, maxOutput: 48000 },
-    "glm-5.0-turbo":      { reasoning: true, thinkingFormat: "openai", thinkingCanDisable: false, contextWindow: 200000, maxOutput: 48000 },
-    "glm-5v-turbo":       { vision: true, reasoning: true, thinkingFormat: "openai", thinkingCanDisable: false, contextWindow: 200000, maxOutput: 38000 },
-    "glm-4.7":            { reasoning: true, thinkingFormat: "openai", contextWindow: 200000, maxOutput: 48000 },
-    "minimax-m3":         { vision: true, reasoning: true, thinkingFormat: "openai", thinkingCanDisable: false, contextWindow: 512000, maxOutput: 48000 },
-    "minimax-m2.7":       { vision: true, reasoning: true, thinkingFormat: "openai", thinkingCanDisable: false, contextWindow: 200000, maxOutput: 48000 },
-    "kimi-k2.7":          { vision: true, reasoning: true, thinkingFormat: "openai", thinkingCanDisable: false, contextWindow: 256000, maxOutput: 32000 },
-    "kimi-k2.6":          { vision: true, reasoning: true, thinkingFormat: "openai", thinkingCanDisable: false, contextWindow: 256000, maxOutput: 32000 },
-    "kimi-k2.5":          { vision: true, reasoning: true, thinkingFormat: "openai", thinkingCanDisable: false, contextWindow: 164000, maxOutput: 32000 },
-    "hy3-preview":        { vision: true, reasoning: true, thinkingFormat: "openai", thinkingCanDisable: false, contextWindow: 192000, maxOutput: 64000 },
-    "deepseek-v4-pro":    { vision: true, reasoning: true, thinkingFormat: "openai", thinkingCanDisable: false, contextWindow: 1000000, maxOutput: 50000 },
-    "deepseek-v4-flash":  { vision: true, reasoning: true, thinkingFormat: "openai", thinkingCanDisable: false, contextWindow: 1000000, maxOutput: 50000 },
-    "deepseek-v3-2-volc": { reasoning: true, thinkingFormat: "openai", thinkingCanDisable: false, contextWindow: 96000, maxOutput: 32000 },
-  },
-};
-
-/**
- * Pattern fallback — glob (* = wildcard), matched case-insensitively and
- * anchored (^...$) so a pattern must match the full model id. ORDER MATTERS:
- * vision/specific variants first, text-only/generic families last, to avoid
- * a broad family pattern swallowing an exception (e.g. glm-4.6v vs glm-5).
- */
-export const PATTERN_CAPABILITIES: Array<{ pattern: string; caps: Partial<Caps> }> = [
-  // ── Claude (4.6+ = adaptive thinking; older/haiku = budget) ──────
-  { pattern: "*claude*opus-4.6*",   caps: { vision: true, reasoning: true, search: true, thinkingFormat: "claude-adaptive" } },
-  { pattern: "*claude*opus-4.7*",   caps: { vision: true, reasoning: true, search: true, thinkingFormat: "claude-adaptive" } },
-  { pattern: "*claude*opus-4.8*",   caps: { vision: true, reasoning: true, search: true, thinkingFormat: "claude-adaptive" } },
-  { pattern: "*claude*sonnet-4.6*", caps: { vision: true, reasoning: true, search: true, thinkingFormat: "claude-adaptive" } },
-  { pattern: "*claude*sonnet-4.7*", caps: { vision: true, reasoning: true, search: true, thinkingFormat: "claude-adaptive" } },
-  { pattern: "*claude*haiku*",  caps: { vision: true, reasoning: true, search: true, thinkingFormat: "claude-budget" } },
-  { pattern: "*claude*opus*",   caps: { vision: true, reasoning: true, search: true, thinkingFormat: "claude-budget" } },
-  { pattern: "*claude*sonnet*", caps: { vision: true, reasoning: true, search: true, thinkingFormat: "claude-budget" } },
-  { pattern: "*claude*fable*",  caps: { vision: true, reasoning: true, search: true, thinkingFormat: "claude-budget", contextWindow: 1000000, maxOutput: 128000 } },
-  { pattern: "*claude*mythos*", caps: { vision: true, reasoning: true, search: true, thinkingFormat: "claude-budget", contextWindow: 1000000, maxOutput: 128000 } },
-  { pattern: "*claude-3*",      caps: { vision: true } },
-  { pattern: "*claude*",        caps: { vision: true, reasoning: true, search: true, thinkingFormat: "claude-budget" } },
-
-  // ── Gemini (all 2.0+ multimodal + google_search grounding, 1M ctx) ─
-  { pattern: "*gemini*image*",  caps: { vision: true, imageOutput: true, contextWindow: 1048576 } },
-  { pattern: "*gemini-3*pro*",  caps: { vision: true, audioInput: true, videoInput: true, reasoning: true, search: true, thinkingFormat: "gemini-level", thinkingCanDisable: false, contextWindow: 1048576, maxOutput: 65535 } },
-  { pattern: "*gemini-3*",      caps: { vision: true, audioInput: true, videoInput: true, reasoning: true, search: true, thinkingFormat: "gemini-level", thinkingCanDisable: false, contextWindow: 1048576, maxOutput: 65536 } },
-  { pattern: "*gemini-2.5*",    caps: { vision: true, audioInput: true, videoInput: true, reasoning: true, search: true, thinkingFormat: "gemini-budget", thinkingRange: { min: 0, max: 24576 }, contextWindow: 1048576, maxOutput: 65536 } },
-  { pattern: "*gemini-2*",      caps: { vision: true, audioInput: true, videoInput: true, search: true, contextWindow: 1048576, maxOutput: 65536 } },
-  { pattern: "*gemini*",        caps: { vision: true, search: true, contextWindow: 1048576 } },
-  { pattern: "*gemma*",         caps: { vision: true, contextWindow: 128000 } },
-  { pattern: "*nanobanana*",    caps: { vision: true, imageOutput: true } },
-
-  // ── OpenAI GPT-5.x (vision + thinking + web search) ──────────────
-  { pattern: "*gpt-5*image*",   caps: { imageOutput: true } },
-  { pattern: "*gpt-5*codex*",   caps: { reasoning: true, search: true, thinkingFormat: "openai", contextWindow: 400000, maxOutput: 128000 } },
-  { pattern: "*gpt-5*",         caps: { vision: true, reasoning: true, search: true, thinkingFormat: "openai", contextWindow: 400000, maxOutput: 128000 } },
-  { pattern: "*gpt-4o*",        caps: { vision: true, search: true, contextWindow: 128000, maxOutput: 16384 } },
-  { pattern: "*gpt-4.1*",       caps: { vision: true, contextWindow: 1000000, maxOutput: 32768 } },
-  { pattern: "*gpt-4-turbo*",   caps: { vision: true, contextWindow: 128000 } },
-  { pattern: "*gpt-4*",         caps: { contextWindow: 128000 } },
-  { pattern: "*gpt-3.5*",       caps: { contextWindow: 16385, maxOutput: 4096 } },
-  { pattern: "*gpt-oss*",       caps: { reasoning: true, thinkingFormat: "openai", contextWindow: 128000 } },
-
-  // ── OpenAI o-series (reasoning, vision) ──────────────────────────
-  { pattern: "*o1-mini*",       caps: { reasoning: true, thinkingFormat: "openai", contextWindow: 128000 } },
-  { pattern: "*o1*",            caps: { vision: true, reasoning: true, thinkingFormat: "openai", contextWindow: 200000, maxOutput: 100000 } },
-  { pattern: "*o3*",            caps: { vision: true, reasoning: true, thinkingFormat: "openai", contextWindow: 200000, maxOutput: 100000 } },
-  { pattern: "*o4*",            caps: { vision: true, reasoning: true, thinkingFormat: "openai", contextWindow: 200000, maxOutput: 100000 } },
-
-  // ── Grok (vision + Live Search) ──────────────────────────────────
-  { pattern: "*grok*image*",    caps: { imageOutput: true } },
-  { pattern: "*grok-code*",     caps: { reasoning: true, thinkingFormat: "openai", contextWindow: 256000 } },
-  { pattern: "*grok-4*",        caps: { vision: true, reasoning: true, search: true, thinkingFormat: "openai", contextWindow: 256000 } },
-  { pattern: "*grok-3*",        caps: { vision: true, reasoning: true, search: true, thinkingFormat: "openai", contextWindow: 131072 } },
-  { pattern: "*grok*",          caps: { vision: true, reasoning: true, search: true, thinkingFormat: "openai", contextWindow: 256000 } },
-
-  // ── Qwen (enable_thinking + thinking_budget; QwQ = thinking-only) ─
-  { pattern: "*qwen*vl*",       caps: { vision: true, reasoning: true, thinkingFormat: "qwen", contextWindow: 262144 } },
-  { pattern: "*qwen*max*",      caps: { vision: true, reasoning: true, thinkingFormat: "qwen", contextWindow: 1000000, maxOutput: 65536 } },
-  { pattern: "*qwen*plus*",     caps: { vision: true, reasoning: true, thinkingFormat: "qwen", contextWindow: 1000000, maxOutput: 65536 } },
-  { pattern: "*qwen*235b*",     caps: { reasoning: true, thinkingFormat: "qwen", contextWindow: 262144 } },
-  { pattern: "*qwen*coder*",    caps: { reasoning: true, thinkingFormat: "qwen", contextWindow: 1000000 } },
-  { pattern: "*qwq*",           caps: { reasoning: true, thinkingFormat: "qwen", thinkingCanDisable: false, contextWindow: 131072 } },
-  { pattern: "*qwen*",          caps: { reasoning: true, thinkingFormat: "qwen", contextWindow: 262144 } },
-
-  // ── Kimi (enabled→reasoning_effort; K2.7-code cannot disable) ─────
-  { pattern: "*kimi*k2.7*code*", caps: { vision: true, reasoning: true, thinkingFormat: "kimi", thinkingCanDisable: false, contextWindow: 262144, maxOutput: 262144 } },
-  { pattern: "*kimi*k2*",       caps: { vision: true, reasoning: true, thinkingFormat: "kimi", contextWindow: 262144, maxOutput: 262144 } },
-  { pattern: "*kimi*",          caps: { reasoning: true, thinkingFormat: "kimi", contextWindow: 262144 } },
-
-  // ── GLM / Z.ai (thinking.enabled; disable via enable_thinking:false) ─
-  { pattern: "*glm-5*",         caps: { reasoning: true, thinkingFormat: "zai", contextWindow: 200000, maxOutput: 128000 } },
-  { pattern: "*glm-4.7*",       caps: { reasoning: true, thinkingFormat: "zai", contextWindow: 200000, maxOutput: 128000 } },
-  { pattern: "*glm-4*",         caps: { reasoning: true, thinkingFormat: "zai", contextWindow: 200000 } },
-  { pattern: "*glm*",           caps: { reasoning: true, thinkingFormat: "zai", contextWindow: 200000 } },
-
-  // ── DeepSeek (thinking.enabled + reasoning_effort; r1 = thinking-only) ─
-  { pattern: "*deepseek-v4*",   caps: { reasoning: true, thinkingFormat: "deepseek", contextWindow: 1000000, maxOutput: 384000 } },
-  { pattern: "*reasoner*",      caps: { reasoning: true, thinkingFormat: "deepseek", thinkingCanDisable: false, contextWindow: 128000 } },
-  { pattern: "*deepseek-r*",    caps: { reasoning: true, thinkingFormat: "deepseek", thinkingCanDisable: false, contextWindow: 128000 } },
-  { pattern: "*deepseek-chat*", caps: { contextWindow: 128000 } },
-  { pattern: "*deepseek*",      caps: { reasoning: true, thinkingFormat: "deepseek", contextWindow: 128000 } },
-
-  // ── MiniMax (M3 = adaptive; M2.x cannot disable) ─────────────────
-  { pattern: "*minimax*image*", caps: { imageOutput: true } },
-  { pattern: "*minimax-m3*",    caps: { vision: true, reasoning: true, thinkingFormat: "minimax", contextWindow: 1048576, maxOutput: 512000 } },
-  { pattern: "*minimax-m2.7*",  caps: { reasoning: true, thinkingFormat: "minimax", thinkingCanDisable: false, contextWindow: 204800, maxOutput: 131072 } },
-  { pattern: "*minimax*",       caps: { reasoning: true, thinkingFormat: "minimax", thinkingCanDisable: false, contextWindow: 200000, maxOutput: 131072 } },
-
-  // ── Xiaomi MiMo (vision, 1M / 262K ctx) ──────────────────────────
-  { pattern: "*mimo*v2.5*",     caps: { vision: true, contextWindow: 1048576, maxOutput: 131072 } },
-  { pattern: "*mimo*omni*",     caps: { vision: true, audioInput: true, contextWindow: 262144, maxOutput: 131072 } },
-  { pattern: "*mimo*",          caps: { vision: true, contextWindow: 262144, maxOutput: 131072 } },
-
-  // ── Llama (4 = vision/1M; 3.x = text-only/128K) ──────────────────
-  { pattern: "*llama-4*",       caps: { vision: true, contextWindow: 1000000 } },
-  { pattern: "*llama*",         caps: { contextWindow: 128000 } },
-
-  // ── Mistral (Large 3 = vision/256K; codestral text) ──────────────
-  { pattern: "*codestral*",     caps: { contextWindow: 256000 } },
-  { pattern: "*mistral-large*", caps: { vision: true, contextWindow: 256000 } },
-  { pattern: "*mistral*",       caps: { contextWindow: 128000 } },
-
-  // ── Cohere (Command A Vision = vision; others text) ──────────────
-  { pattern: "*command-a-vision*", caps: { vision: true, contextWindow: 128000 } },
-  { pattern: "*command*",       caps: { contextWindow: 128000 } },
-
-  // ── Perplexity (web search native) ───────────────────────────────
-  { pattern: "*sonar*",         caps: { search: true, contextWindow: 128000 } },
-  { pattern: "*pplx*",          caps: { search: true, contextWindow: 128000 } },
-  { pattern: "*perplexity*",    caps: { search: true, contextWindow: 128000 } },
-
-  // ── Others ───────────────────────────────────────────────────────
-  { pattern: "*hunyuan*",       caps: { reasoning: true, thinkingFormat: "hunyuan", contextWindow: 262144, maxOutput: 262144 } },
-  { pattern: "hy3*",            caps: { reasoning: true, thinkingFormat: "hunyuan", contextWindow: 262144, maxOutput: 262144 } },
-  { pattern: "*step-*",         caps: { reasoning: true, thinkingFormat: "step", contextWindow: 128000 } },
-  { pattern: "*nemotron*",      caps: { reasoning: true, contextWindow: 128000 } },
-  { pattern: "*ling-*",         caps: { reasoning: true, contextWindow: 128000 } },
-];
-
-/**
- * Resolve capabilities for a model using the 4-step fallback chain,
- * merged over DEFAULT_CAPABILITIES so the result is always complete.
- *
- * @param {string} provider
- * @param {string} model
- * @returns {object} full capabilities object
- */
-/**
- * models.dev-style modalities for a model ref, derived from its capabilities.
- * Ref may be a `provider/model` (vendor prefix tolerated) or a bare combo alias.
- * Mirrors what opencode/aigetwey store per model in opencode.json.
- */
-export function modalitiesForModel(ref: string): { input: string[]; output: string[] } {
-  const slash = ref.indexOf("/");
-  const provider = slash > 0 ? ref.slice(0, slash) : null;
-  const model = slash > 0 ? ref.slice(slash + 1) : ref;
-  const c = getCapabilitiesForModel(provider, model);
-  const input = ["text"];
-  if (c.vision) input.push("image");
-  if (c.pdf) input.push("pdf");
-  if (c.audioInput) input.push("audio");
-  if (c.videoInput) input.push("video");
-  const output = ["text"];
-  if (c.imageOutput) output.push("image");
-  if (c.audioOutput) output.push("audio");
-  return { input, output };
-}
-
-export function getCapabilitiesForModel(provider: string | null, model: string): Caps {
-  if (!model) return { ...DEFAULT_CAPABILITIES };
-
-  // 1. Provider-specific override
-  if (provider && PROVIDER_CAPABILITIES[provider]?.[model]) {
-    return { ...DEFAULT_CAPABILITIES, ...PROVIDER_CAPABILITIES[provider][model] };
-  }
-
-  // 2. Canonical exact (strip vendor prefix: "anthropic/claude-opus-4.7" -> "claude-opus-4.7")
-  const baseModel = (model.includes("/") ? model.split("/").pop() : model) ?? model;
-  if (MODEL_CAPABILITIES[baseModel]) return { ...DEFAULT_CAPABILITIES, ...MODEL_CAPABILITIES[baseModel] };
-  if (MODEL_CAPABILITIES[model]) return { ...DEFAULT_CAPABILITIES, ...MODEL_CAPABILITIES[model] };
-
-  // 3. Pattern match (first match wins)
-  for (const { pattern, caps } of PATTERN_CAPABILITIES) {
-    if (matchPattern(pattern, baseModel) || matchPattern(pattern, model)) {
-      return { ...DEFAULT_CAPABILITIES, ...caps };
-    }
-  }
-
-  // 4. Floor
-  return { ...DEFAULT_CAPABILITIES };
-}

package/dashboard/src/lib/cliTools.ts

@@ -1,77 +0,0 @@
-/**
- * CLI tool setup definitions. Each tool talks to the gateway in either OpenAI or
- * Anthropic wire format; the gateway exposes both on the same port, so setup is
- * just pointing the tool's base_url + key at us. `env` builders take the live
- * gateway base URL + a gateway key and return ready-to-copy environment lines.
- *
- * `slots` are the model names this tool calls. Because our gateway routes by the
- * combo alias (the alias IS the model name), the detail page checks whether a
- * combo with each slot's alias exists, and prompts to create the missing ones.
- */
-export type ToolFormat = "openai" | "anthropic";
-
-export interface CliTool {
-  id: string;
-  name: string;
-  /** Material Symbols ligature shown on the tool card + detail header. */
-  icon: string;
-  format: ToolFormat;
-  blurb: string;
-  /** true when the dashboard can detect + write this tool's local config file. */
-  autoConfig?: boolean;
-  /** one-line install command, when the tool ships via a package manager. */
-  install?: string;
-  /** model names the tool sends — pair each with a combo of the same name. */
-  slots: { label: string; alias: string }[];
-  /** environment variables to set, given the gateway base + a key. */
-  env: (base: string, key: string) => Array<{ name: string; value: string }>;
-  /** extra free-form steps shown on the detail page. */
-  steps: string[];
-}
-
-const KEY = (k: string) => k || "<your-gateway-key>";
-
-export const CLI_TOOLS: CliTool[] = [
-  {
-    id: "claude-code",
-    name: "Claude Code",
-    icon: "smart_toy",
-    format: "anthropic",
-    blurb: "Anthropic CLI. Point its base URL + key at the gateway.",
-    autoConfig: true,
-    install: "npm i -g @anthropic-ai/claude-code",
-    slots: [
-      { label: "Opus · heavy", alias: "claude-opus-4-1" },
-      { label: "Sonnet · default", alias: "claude-sonnet-4-6" },
-      { label: "Haiku · fast", alias: "claude-haiku-4-5" },
-    ],
-    env: (base, key) => [
-      { name: "ANTHROPIC_BASE_URL", value: base },
-      { name: "ANTHROPIC_API_KEY", value: KEY(key) },
-    ],
-    steps: [
-      "Export the two variables in the shell you run `claude` from.",
-      "Create a combo named like each slot above so Claude Code's model ids resolve.",
-      "The gateway translates Anthropic ↔ provider format, so any provider works behind it.",
-    ],
-  },
-  {
-    id: "opencode",
-    name: "opencode",
-    icon: "code_blocks",
-    format: "openai",
-    blurb: "OpenAI-compatible provider. Set base_url to /v1.",
-    autoConfig: true,
-    install: "curl -fsSL https://opencode.ai/install | bash",
-    slots: [{ label: "Model", alias: "gpt-5" }],
-    env: (base, key) => [
-      { name: "OPENAI_BASE_URL", value: `${base}/v1` },
-      { name: "OPENAI_API_KEY", value: KEY(key) },
-    ],
-    steps: ["Add an OpenAI-compatible provider with the gateway /v1 base URL.", "Pick a combo alias as the model."],
-  },
-];
-
-export function toolById(id: string): CliTool | undefined {
-  return CLI_TOOLS.find((t) => t.id === id);
-}

package/dashboard/src/lib/client.ts

@@ -1,210 +0,0 @@
-"use client";
-
-/**
- * Browser-side admin client. Talks to the gateway THROUGH the same-origin proxy
- * (`/api/gw/admin/...`), which injects the admin Bearer server-side — so the
- * password never reaches the browser. Server components read initial data via
- * lib/gateway.ts directly; client components mutate through here.
- */
-import type {
-  BudgetStatus,
-  ConfigReply,
-  EndpointPayload,
-  HeadroomStatusReply,
-  InjectLevel,
-  KeyUsageRow,
-  ModelsPayload,
-  PingResult,
-  PricingPayload,
-  ProviderSnapshot,
-  WireFormat,
-} from "./gateway";
-
-export interface ApiResult<T> {
-  ok: boolean;
-  status: number;
-  data: T | null;
-  error?: string;
-}
-
-async function api<T>(method: string, path: string, body?: unknown): Promise<ApiResult<T>> {
-  let res: Response;
-  try {
-    res = await fetch(`/api/gw${path}`, {
-      method,
-      headers: body !== undefined ? { "content-type": "application/json" } : {},
-      body: body !== undefined ? JSON.stringify(body) : undefined,
-    });
-  } catch (e) {
-    return { ok: false, status: 0, data: null, error: (e as Error).message };
-  }
-  const text = await res.text();
-  let data: T | null = null;
-  try {
-    data = text ? (JSON.parse(text) as T) : null;
-  } catch {
-    data = null;
-  }
-  if (!res.ok) {
-    const err = (data as { error?: string } | null)?.error ?? `request failed (${res.status})`;
-    return { ok: false, status: res.status, data, error: err };
-  }
-  return { ok: true, status: res.status, data };
-}
-
-export const adminApi = {
-  providers: () => api<{ providers: ProviderSnapshot[] }>("GET", "/admin/providers"),
-  budgets: () => api<{ budgets: BudgetStatus[] }>("GET", "/admin/budgets"),
-  models: () => api<ModelsPayload>("GET", "/admin/models"),
-  keys: () => api<Array<{ fingerprint: string; name: string; masked: string }>>("GET", "/admin/keys"),
-  keysUsage: () => api<{ keys: KeyUsageRow[] }>("GET", "/admin/keys/usage"),
-
-  setBudget: (body: {
-    scope: { type: "global" } | { type: "provider"; id: string } | { type: "model"; id: string } | { type: "key"; id: string };
-    unit: "usd" | "tokens";
-    limit: number;
-    window: "5h" | "24h" | "7day" | "30day";
-    alert_at?: number;
-    note?: string;
-  }) => api<ConfigReply>("PUT", "/admin/budgets", body),
-  clearBudget: (key: string) => api<ConfigReply>("DELETE", `/admin/budgets/${encodeURIComponent(key)}`),
-
-  addProvider: (p: {
-    id: string;
-    name?: string;
-    format: WireFormat;
-    base_url: string;
-    api_key?: string;
-    free?: boolean;
-    auto_models?: boolean;
-    service_account?: string;
-  }) => api<ConfigReply>("POST", "/admin/providers", p),
-  editProvider: (id: string, patch: { base_url?: string; format?: WireFormat; name?: string }) =>
-    api<ConfigReply>("PUT", `/admin/providers/${encodeURIComponent(id)}`, patch),
-  renameProvider: (id: string, newId: string) =>
-    api<ConfigReply>("PUT", `/admin/providers/${encodeURIComponent(id)}/rename`, { id: newId }),
-  removeProvider: (id: string) => api<ConfigReply>("DELETE", `/admin/providers/${encodeURIComponent(id)}`),
-  addKey: (id: string, key: string, name?: string) =>
-    api<ConfigReply>("POST", `/admin/providers/${encodeURIComponent(id)}/keys`, { key, name }),
-  editKey: (id: string, index: number, patch: { key?: string; name?: string }) =>
-    api<ConfigReply>("PUT", `/admin/providers/${encodeURIComponent(id)}/keys/${index}`, patch),
-  removeKey: (id: string, index: number) =>
-    api<ConfigReply>("DELETE", `/admin/providers/${encodeURIComponent(id)}/keys/${index}`),
-  revealKey: (id: string, index: number) =>
-    api<{ key: string }>("GET", `/admin/providers/${encodeURIComponent(id)}/keys/${index}/reveal`),
-  addModel: (id: string, model: string, price?: { price_in?: number; price_out?: number }) =>
-    api<ConfigReply>("POST", `/admin/providers/${encodeURIComponent(id)}/models`, { model, ...price }),
-  addModels: (id: string, models: string[]) =>
-    api<ConfigReply>("POST", `/admin/providers/${encodeURIComponent(id)}/models`, { models }),
-  // model id can hold a slash (provider/model); send it as a query param so the
-  // proxy's path split doesn't mangle it.
-  removeModel: (id: string, model: string) =>
-    api<ConfigReply>("DELETE", `/admin/providers/${encodeURIComponent(id)}/models?model=${encodeURIComponent(model)}`),
-  clearModels: (id: string) => api<ConfigReply>("DELETE", `/admin/providers/${encodeURIComponent(id)}/models`),
-  pricing: () => api<PricingPayload>("GET", "/admin/pricing"),
-  setModelPrice: (id: string, model: string, price: { price_in?: number | null; price_out?: number | null }) =>
-    api<ConfigReply>("PUT", `/admin/providers/${encodeURIComponent(id)}/models/price`, { model, ...price }),
-  testProvider: (id: string) => api<PingResult>("POST", `/admin/providers/${encodeURIComponent(id)}/test`),
-  testKey: (id: string, index: number) =>
-    api<PingResult>("POST", `/admin/providers/${encodeURIComponent(id)}/keys/${index}/test`),
-  testModel: (id: string, model: string) =>
-    api<{ ok: boolean; status?: number; error?: string }>(
-      "POST",
-      `/admin/providers/${encodeURIComponent(id)}/models/test?model=${encodeURIComponent(model)}`,
-    ),
-  validateProvider: (b: { format: WireFormat; base_url: string; api_key?: string }) =>
-    api<PingResult>("POST", "/admin/providers/validate", b),
-  // discover: returns the upstream catalog flagged with which ids are in config.
-  discoverModels: (id: string) =>
-    api<{ ok: boolean; models: Array<{ id: string; added: boolean }> }>(
-      "POST",
-      `/admin/providers/${encodeURIComponent(id)}/connect`,
-    ),
-
-  reorderProvider: (from: number, to: number) =>
-    api<ConfigReply>("PUT", "/admin/providers/reorder", { from, to }),
-  reorderKey: (id: string, from: number, to: number) =>
-    api<ConfigReply>("PUT", `/admin/providers/${encodeURIComponent(id)}/keys/reorder`, { from, to }),
-  toggleKey: (id: string, index: number, enabled: boolean) =>
-    api<ConfigReply>("PUT", `/admin/providers/${encodeURIComponent(id)}/keys/${index}/toggle`, { enabled }),
-  setProviderStrategy: (id: string, strategy: "fallback" | "round-robin" | null, sticky?: number) =>
-    api<ConfigReply>("PUT", `/admin/providers/${encodeURIComponent(id)}/strategy`, { strategy, sticky }),
-  setProviderDisabled: (id: string, disabled: boolean) =>
-    api<ConfigReply>("PUT", `/admin/providers/${encodeURIComponent(id)}/disabled`, { disabled }),
-
-  setRoute: (
-    alias: string,
-    body: { target: string[]; model?: string | string[]; strategy?: "fallback" | "round-robin"; price_in?: number; price_out?: number },
-  ) => api<ConfigReply>("PUT", `/admin/routes/${encodeURIComponent(alias)}`, body),
-  removeRoute: (alias: string) => api<ConfigReply>("DELETE", `/admin/routes/${encodeURIComponent(alias)}`),
-
-  endpoint: () => api<EndpointPayload>("GET", "/admin/endpoint"),
-  setRtk: (enabled: boolean) => api<ConfigReply>("PUT", "/admin/endpoint/rtk", { enabled }),
-  setCaveman: (level: InjectLevel) => api<ConfigReply>("PUT", "/admin/endpoint/caveman", { level }),
-  setPonytail: (level: InjectLevel) => api<ConfigReply>("PUT", "/admin/endpoint/ponytail", { level }),
-  addServerKey: (key: string, name?: string) => api<ConfigReply>("POST", "/admin/endpoint/keys", { key, name }),
-  editServerKey: (index: number, name: string) => api<ConfigReply>("PUT", `/admin/endpoint/keys/${index}`, { name }),
-  setServerKeyScope: (index: number, body: { models?: string[]; rpm?: number | null; expires?: number | null }) =>
-    api<ConfigReply>("PUT", `/admin/endpoint/keys/${index}/scope`, body),
-  removeServerKey: (index: number) => api<ConfigReply>("DELETE", `/admin/endpoint/keys/${index}`),
-  revealServerKey: (index: number) => api<{ key: string }>("GET", `/admin/endpoint/keys/${index}/reveal`),
-
-  setHeadroom: (patch: { enabled?: boolean; url?: string; compress_user_messages?: boolean }) =>
-    api<ConfigReply>("PUT", "/admin/endpoint/headroom", patch),
-  headroomStatus: () => api<HeadroomStatusReply>("GET", "/admin/headroom/status"),
-  headroomStart: () => api<{ success?: boolean; pid?: number; alreadyRunning?: boolean }>("POST", "/admin/headroom/start"),
-  headroomStop: () => api<{ stopped: boolean; reason?: string; pid?: number }>("POST", "/admin/headroom/stop"),
-
-  putConfig: (text: string) => api<{ ok: boolean }>("PUT", "/admin/config", { text }),
-
-  version: () => api<{ current: string; latest: string | null; updateAvailable: boolean }>("GET", "/admin/version"),
-  shutdown: () => api<{ ok: boolean; message: string }>("POST", "/admin/shutdown"),
-};
-
-// Local CLI-tool detection/auto-config. These hit the dashboard's OWN server
-// routes (/api/cli-detect/*), not the gateway proxy — they read/write the tool's
-// config file on this machine. Session-gated by middleware like everything else.
-export interface CliStatus {
-  auto: boolean;
-  installed: boolean;
-  configured?: boolean;
-  path?: string;
-  baseUrl?: string | null;
-  models?: string[];
-  activeModel?: string | null;
-  // claude returns its three slot defaults instead of a flat list
-  modelSlots?: { opus?: string | null; sonnet?: string | null; haiku?: string | null };
-}
-
-async function appApi<T>(method: string, url: string, body?: unknown): Promise<ApiResult<T>> {
-  let res: Response;
-  try {
-    res = await fetch(url, {
-      method,
-      headers: body !== undefined ? { "content-type": "application/json" } : {},
-      body: body !== undefined ? JSON.stringify(body) : undefined,
-    });
-  } catch (e) {
-    return { ok: false, status: 0, data: null, error: (e as Error).message };
-  }
-  const text = await res.text();
-  const data = text ? (JSON.parse(text) as T) : null;
-  if (!res.ok) {
-    const err = (data as { error?: string } | null)?.error ?? `request failed (${res.status})`;
-    return { ok: false, status: res.status, data, error: err };
-  }
-  return { ok: true, status: res.status, data };
-}
-
-/** Admin account actions that hit the dashboard's OWN routes (not the gw proxy). */
-export const account = {
-  changePassword: (current: string, next: string) =>
-    appApi<{ ok: boolean }>("POST", "/api/password", { current, next }),
-};
-
-export const cliConfig = {
-  status: (tool: string) => appApi<CliStatus>("GET", `/api/cli-detect/${encodeURIComponent(tool)}`),
-  apply: (tool: string, body: { base: string; key?: string; models?: string[] | Record<string, string>; active?: string }) =>
-    appApi<{ success?: boolean; path?: string }>("POST", `/api/cli-detect/${encodeURIComponent(tool)}`, body),
-  reset: (tool: string) => appApi<{ success?: boolean }>("DELETE", `/api/cli-detect/${encodeURIComponent(tool)}`),
-};