[{"_1":2,"_3":-5,"_4":-5},"loaderData",{"_5":6,"_7":8,"_9":10},"actionData","errors","root",{"_154":155,"_156":157},"routes/_app.x.$handle.$name",{"_94":93,"_32":45,"_103":84,"_27":104,"_13":14,"_105":-5,"_11":106,"_107":108,"_109":110,"_111":112,"_113":114,"_20":-5,"_115":-5,"_116":-5,"_117":25,"_118":119,"_120":121,"_122":-5,"_123":-5,"_124":125},"routes/_app.x.$handle.$name.branch.$branch.version.$version.diff",{"_11":12,"_13":14,"_15":16,"_17":-5,"_18":-5},"project",{"_43":44,"_32":45,"_38":46,"_47":-5,"_48":25,"_49":50,"_51":50,"_52":-5,"_53":-5,"_18":54,"_55":-5,"_56":25,"_57":-5,"_58":-5,"_59":60,"_61":62,"_63":-5,"_64":-5,"_65":-5,"_66":-5,"_67":68,"_69":68,"_70":71,"_72":71,"_73":71,"_74":68,"_75":68,"_76":77,"_78":79,"_80":81,"_82":25},"version",16,"changes",[19],"aiChatInfo","userId",{"_20":21,"_22":23,"_24":25,"_26":25,"_27":28,"_29":28},"file",{"_30":31,"_32":28,"_33":34,"_35":37,"_38":39,"_40":14,"_41":42,"_24":25},"previous",{"_30":31,"_32":28,"_33":34,"_35":36},"deleted",false,"deletedHere","path","proxy.ts","latestPath","fileId","45c3c7dc-60a9-11f1-a1a4-1607ee4eb77e","name","fileType","http","content","/**\n * FreeAI Proxy — OpenAI + Anthropic compatible, round-robins through relay pool\n *\n * OpenAI: GET /v1/models\n * POST /v1/chat/completions\n * Anthropic: POST /v1/messages\n */\n\nconst UPSTREAM = \"https://opencode.ai/zen/v1\";\nconst ZEN_KEY = \"public\";\nconst MODELS_DEV_URL = \"https://models.dev/api.json\";\nconst UPSTREAM_URL = `${UPSTREAM}/chat/completions`;\n\nconst RELAYS = [\n \"https://joddabod--2cb9984260aa11f1af5d1607ee4eb77e.web.val.run\",\n \"https://joddabod--2ce78ac260aa11f1bcfe1607ee4eb77e.web.val.run\",\n \"https://joddabod--2d4384c660aa11f195321607ee4eb77e.web.val.run\",\n \"https://joddabod--2d845b0460aa11f1b5531607ee4eb77e.web.val.run\",\n \"https://joddabod--2db5872e60aa11f1b7d71607ee4eb77e.web.val.run\",\n];\n\nlet relayIndex = 0;\n\nfunction nextRelay(): string {\n const url = RELAYS[relayIndex % RELAYS.length];\n relayIndex = (relayIndex + 1) >>> 0;\n return url;\n}\n\n// ─── CORS helpers ──────────────────────────────────────────────────────────────\n\nfunction corsHeaders(): Record {\n return {\n \"access-control-allow-origin\": \"*\",\n \"access-control-allow-methods\": \"GET, POST, OPTIONS\",\n \"access-control-allow-headers\": \"Content-Type, Authorization, x-api-key, anthropic-version\",\n \"access-control-max-age\": \"86400\",\n };\n}\n\nfunction json(data: unknown, status = 200): Response {\n return new Response(JSON.stringify(data), {\n status,\n headers: {\n \"content-type\": \"application/json\",\n ...corsHeaders(),\n },\n });\n}\n\nfunction apiError(message: string, status = 400): Response {\n return json({ error: { message, type: \"invalid_request_error\", code: null } }, status);\n}\n\n// ─── Free model list (cached) ──────────────────────────────────────────────────\n\nconst FALLBACK_MODELS = [\n { id: \"deepseek-v4-flash-free\", object: \"model\", created: 1745452800, owned_by: \"opencode-zen\", name: \"DeepSeek V4 Flash\", context_length: 200_000, capabilities: { reasoning: true, vision: false, tool_call: true }, pricing: { input: 0, output: 0 } },\n { id: \"big-pickle\", object: \"model\", created: 1729123200, owned_by: \"opencode-zen\", name: \"Big Pickle\", context_length: 200_000, capabilities: { reasoning: true, vision: false, tool_call: true }, pricing: { input: 0, output: 0 } },\n { id: \"mimo-v2.5-free\", object: \"model\", created: 1745452800, owned_by: \"opencode-zen\", name: \"MiMo 2.5\", context_length: 200_000, capabilities: { reasoning: true, vision: true, tool_call: true }, pricing: { input: 0, output: 0 } },\n];\n\nlet _modelsCache: typeof FALLBACK_MODELS | null = null;\nlet _modelsCacheAt = 0;\nconst CACHE_TTL = 5 * 60 * 1000;\n\nasync function getFreeModels() {\n const now = Date.now();\n if (_modelsCache && now - _modelsCacheAt < CACHE_TTL) return _modelsCache;\n\n try {\n const res = await fetch(MODELS_DEV_URL, { signal: AbortSignal.timeout(6000) });\n if (!res.ok) throw new Error(\"models.dev non-200\");\n const data = await res.json();\n const models = Object.values(data?.opencode?.models ?? {})\n .filter((m: any) => m.status !== \"deprecated\" && (m.cost?.input ?? 1) === 0)\n .map(toOpenAIModel);\n\n if (models.length) {\n _modelsCache = models;\n _modelsCacheAt = now;\n }\n return _modelsCache ?? FALLBACK_MODELS;\n } catch {\n return _modelsCache ?? FALLBACK_MODELS;\n }\n}\n\nfunction toOpenAIModel(m: any) {\n return {\n id: m.id,\n object: \"model\",\n created: Math.floor(new Date(m.release_date || \"2025-01-01\").getTime() / 1000),\n owned_by: \"opencode-zen\",\n name: m.name,\n context_length: m.limit?.context ?? 200_000,\n capabilities: {\n reasoning: m.reasoning ?? false,\n vision: m.attachment ?? m.modalities?.input?.some((i: string) => [\"image\", \"video\", \"audio\"].includes(i)) ?? false,\n tool_call: m.tool_call ?? true,\n },\n pricing: { input: 0, output: 0 },\n };\n}\n\n// ─── Relay fetch helper ────────────────────────────────────────────────────────\n\nasync function relayFetch(body: unknown): Promise<{ response: Response; relayUrl: string }> {\n const relayUrl = nextRelay();\n const bodyJson = JSON.stringify(body);\n const response = await fetch(relayUrl, {\n method: \"POST\",\n headers: {\n \"content-type\": \"application/json\",\n \"x-relay-target\": UPSTREAM_URL,\n authorization: `Bearer ${ZEN_KEY}`,\n },\n body: bodyJson,\n });\n return { response, relayUrl };\n}\n\nfunction copyRateLimitHeaders(upstream: Response, target: Record) {\n for (const h of [\n \"x-ratelimit-limit-requests\", \"x-ratelimit-limit-tokens\",\n \"x-ratelimit-remaining-requests\", \"x-ratelimit-remaining-tokens\",\n \"x-ratelimit-reset-requests\", \"x-ratelimit-reset-tokens\",\n ]) {\n const v = upstream.headers.get(h);\n if (v) target[h] = v;\n }\n}\n\n// ─── OpenAI endpoints ──────────────────────────────────────────────────────────\n\nasync function handleModels(): Promise {\n const models = await getFreeModels();\n return json({ object: \"list\", data: models });\n}\n\nasync function handleCompletions(req: Request): Promise {\n let body: any;\n try { body = await req.json(); } catch { return apiError(\"Request body must be valid JSON\"); }\n if (!body.model) return apiError(\"Missing required field: model\");\n if (!body.messages) return apiError(\"Missing required field: messages\");\n\n try {\n const { response: upstream, relayUrl } = await relayFetch(body);\n const resHeaders: Record = { ...corsHeaders() };\n const ct = upstream.headers.get(\"content-type\");\n if (ct) resHeaders[\"content-type\"] = ct;\n copyRateLimitHeaders(upstream, resHeaders);\n return new Response(upstream.body, { status: upstream.status, headers: resHeaders });\n } catch (err: any) {\n return apiError(`Relay failed: ${err.message}`, 502);\n }\n}\n\n// ─── Anthropic endpoint ────────────────────────────────────────────────────────\n\n/** Extract auth token from either OpenAI-style (Authorization) or Anthropic-style (x-api-key) */\nfunction extractBearer(req: Request): string {\n const auth = req.headers.get(\"authorization\") || \"\";\n if (auth.startsWith(\"Bearer \")) return auth.slice(7);\n const key = req.headers.get(\"x-api-key\") || \"\";\n return key;\n}\n\n/** Flatten Anthropic content blocks into a single text string */\nfunction flattenAnthropicContent(content: any): string {\n if (typeof content === \"string\") return content;\n if (Array.isArray(content)) {\n return content\n .filter((b: any) => b.type === \"text\" || b.type === \"image_url\" || b.type === \"image\")\n .map((b: any) => {\n if (b.type === \"text\") return b.text;\n return \"[image]\";\n })\n .join(\"\\n\");\n }\n return String(content);\n}\n\n/** Convert an Anthropic /v1/messages request body to OpenAI /chat/completions format */\nfunction anthropicToOpenAI(body: any): any {\n const openAIMessages: any[] = [];\n\n // System prompt → system message\n if (body.system) {\n if (typeof body.system === \"string\") {\n openAIMessages.push({ role: \"system\", content: body.system });\n } else if (Array.isArray(body.system)) {\n const text = body.system.filter((b: any) => b.type === \"text\").map((b: any) => b.text).join(\"\\n\");\n if (text) openAIMessages.push({ role: \"system\", content: text });\n }\n }\n\n // Messages\n for (const msg of body.messages || []) {\n const content = flattenAnthropicContent(msg.content);\n if (content) {\n openAIMessages.push({ role: msg.role, content });\n }\n }\n\n return {\n model: body.model,\n messages: openAIMessages,\n max_tokens: body.max_tokens ?? 4096,\n temperature: body.temperature,\n top_p: body.top_p,\n stream: body.stream ?? false,\n };\n}\n\n/** Mapping from OpenAI finish_reason to Anthropic stop_reason */\nfunction mapStopReason(finishReason: string | null): string {\n switch (finishReason) {\n case \"stop\": return \"end_turn\";\n case \"length\": return \"max_tokens\";\n case \"content_filter\": return \"content_filtered\";\n case \"tool_calls\": return \"tool_use\";\n default: return \"end_turn\";\n }\n}\n\n/** Convert an OpenAI non-streaming response to Anthropic /v1/messages format */\nfunction openAIToAnthropic(openAIRes: any, model: string): any {\n const choice = openAIRes.choices?.[0] || {};\n const message = choice.message || {};\n const usage = openAIRes.usage || {};\n\n return {\n id: openAIRes.id || `msg_${crypto.randomUUID()}`,\n type: \"message\",\n role: \"assistant\",\n content: [\n {\n type: \"text\",\n text: message.content || \"\",\n },\n ],\n model: model,\n stop_reason: mapStopReason(choice.finish_reason),\n stop_sequence: null,\n usage: {\n input_tokens: usage.prompt_tokens || 0,\n output_tokens: usage.completion_tokens || 0,\n },\n };\n}\n\n/** Translate one OpenAI SSE chunk into zero or more Anthropic SSE strings */\nfunction translateStreamChunk(\n openAIChunk: any,\n model: string,\n state: { messageStarted: boolean; contentStarted: boolean; contentIndex: number },\n): string[] {\n const events: string[] = [];\n const delta = openAIChunk.choices?.[0]?.delta || {};\n const finishReason = openAIChunk.choices?.[0]?.finish_reason || null;\n const usage = openAIChunk.usage || null;\n\n // message_start\n if (!state.messageStarted) {\n state.messageStarted = true;\n events.push(`event: message_start\ndata: ${JSON.stringify({\n type: \"message_start\",\n message: {\n id: `msg_${crypto.randomUUID()}`,\n type: \"message\",\n role: \"assistant\",\n content: [],\n model,\n stop_reason: null,\n stop_sequence: null,\n usage: { input_tokens: 0, output_tokens: 0 },\n },\n})}\n\n`);\n }\n\n // content_block_start\n if (delta.content && !state.contentStarted) {\n state.contentStarted = true;\n events.push(`event: content_block_start\ndata: ${JSON.stringify({\n type: \"content_block_start\",\n index: state.contentIndex,\n content_block: { type: \"text\", text: \"\" },\n})}\n\n`);\n }\n\n // content_block_delta\n if (delta.content) {\n events.push(`event: content_block_delta\ndata: ${JSON.stringify({\n type: \"content_block_delta\",\n index: state.contentIndex,\n delta: { type: \"text_delta\", text: delta.content },\n})}\n\n`);\n }\n\n // reasoning\n if (delta.reasoning_content) {\n events.push(`event: content_block_delta\ndata: ${JSON.stringify({\n type: \"content_block_delta\",\n index: state.contentIndex,\n delta: { type: \"reasoning_delta\", reasoning: delta.reasoning_content },\n})}\n\n`);\n }\n\n // content_block_stop + message_delta on finish\n if (finishReason && state.contentStarted) {\n events.push(`event: content_block_stop\ndata: ${JSON.stringify({ type: \"content_block_stop\", index: state.contentIndex })}\n\n`);\n events.push(`event: message_delta\ndata: ${JSON.stringify({\n type: \"message_delta\",\n delta: { stop_reason: mapStopReason(finishReason), stop_sequence: null },\n usage: usage ? { output_tokens: usage.completion_tokens || 0 } : undefined,\n})}\n\n`);\n }\n\n // message_stop\n if (finishReason) {\n events.push(`event: message_stop\ndata: ${JSON.stringify({ type: \"message_stop\" })}\n\n`);\n }\n\n return events;\n}\n\nasync function handleAnthropicMessages(req: Request): Promise {\n // Read auth token\n const bearer = extractBearer(req);\n\n let body: any;\n try { body = await req.json(); } catch { return apiError(\"Request body must be valid JSON\"); }\n if (!body.model) return apiError(\"Missing required field: model\");\n if (!body.messages) return apiError(\"Missing required field: messages\");\n if (!body.max_tokens) return apiError(\"Missing required field: max_tokens\");\n\n const model = body.model;\n const streaming = body.stream ?? false;\n const openAIBody = anthropicToOpenAI(body);\n\n if (streaming) {\n // Streaming: read the OpenAI stream and translate each chunk\n try {\n const relayUrl = nextRelay();\n const upstream = await fetch(relayUrl, {\n method: \"POST\",\n headers: {\n \"content-type\": \"application/json\",\n \"x-relay-target\": UPSTREAM_URL,\n authorization: `Bearer ${ZEN_KEY}`,\n },\n body: JSON.stringify(openAIBody),\n });\n\n if (!upstream.ok) {\n const errText = await upstream.text().catch(() => \"Upstream error\");\n return json(\n { error: { type: \"api_error\", message: errText.slice(0, 200) } },\n upstream.status,\n );\n }\n\n const { readable, writable } = new TransformStream();\n const writer = writable.getWriter();\n const encoder = new TextEncoder();\n\n const state = { messageStarted: false, contentStarted: false, contentIndex: 0 };\n\n (async () => {\n try {\n const reader = upstream.body!.getReader();\n const decoder = new TextDecoder();\n let buf = \"\";\n\n while (true) {\n const { done, value } = await reader.read();\n if (done) break;\n buf += decoder.decode(value, { stream: true });\n const lines = buf.split(\"\\n\");\n buf = lines.pop() || \"\";\n\n for (const line of lines) {\n const t = line.trim();\n if (!t || t === \"data: [DONE]\" || !t.startsWith(\"data: \")) continue;\n try {\n const chunk = JSON.parse(t.slice(6));\n const events = translateStreamChunk(chunk, model, state);\n for (const evt of events) {\n await writer.write(encoder.encode(evt));\n }\n } catch { /* skip malformed */ }\n }\n }\n // If stream ended but we never got a finish_reason, close cleanly\n if (state.contentStarted) {\n await writer.write(encoder.encode(`event: content_block_stop\\ndata: ${JSON.stringify({ type: \"content_block_stop\", index: 0 })}\\n\\n`));\n await writer.write(encoder.encode(`event: message_delta\\ndata: ${JSON.stringify({ type: \"message_delta\", delta: { stop_reason: \"end_turn\", stop_sequence: null } })}\\n\\n`));\n await writer.write(encoder.encode(`event: message_stop\\ndata: ${JSON.stringify({ type: \"message_stop\" })}\\n\\n`));\n }\n } catch { /* stream interrupted */ }\n await writer.close();\n })();\n\n return new Response(readable, {\n headers: {\n \"content-type\": \"text/event-stream\",\n \"cache-control\": \"no-cache\",\n ...corsHeaders(),\n },\n });\n } catch (err: any) {\n return apiError(`Relay failed: ${err.message}`, 502);\n }\n } else {\n // Non-streaming: proxy and convert response\n try {\n const { response: upstream } = await relayFetch(openAIBody);\n const upstreamData = await upstream.json().catch(() => null);\n\n if (!upstream.ok || !upstreamData) {\n const msg = upstreamData?.error?.message || `Upstream error ${upstream.status}`;\n return json(\n { error: { type: \"api_error\", message: msg } },\n upstream.status,\n );\n }\n\n const anthropicRes = openAIToAnthropic(upstreamData, model);\n const resHeaders: Record = { ...corsHeaders() };\n copyRateLimitHeaders(upstream, resHeaders);\n return json(anthropicRes, upstream.status);\n } catch (err: any) {\n return apiError(`Relay failed: ${err.message}`, 502);\n }\n }\n}\n\n// ─── Router ─────────────────────────────────────────────────────────────────────\n\nexport default async function handler(req: Request): Promise {\n const url = new URL(req.url);\n const method = req.method;\n\n if (method === \"OPTIONS\") {\n return new Response(null, { status: 204, headers: corsHeaders() });\n }\n\n if (method === \"GET\" && url.pathname === \"/v1/models\") {\n return handleModels();\n }\n\n if (method === \"POST\" && url.pathname === \"/v1/chat/completions\") {\n return handleCompletions(req);\n }\n\n // Anthropic-compatible endpoint\n if (method === \"POST\" && url.pathname === \"/v1/messages\") {\n return handleAnthropicMessages(req);\n }\n\n return apiError(\"Not found\", 404);\n}\n","/**\n * FreeAI Proxy — OpenAI + Anthropic compatible, round-robins through relay pool\n *\n * OpenAI: GET /v1/models\n * POST /v1/chat/completions\n * Anthropic: POST /v1/messages\n */\n\nconst UPSTREAM = \"https://opencode.ai/zen/v1\";\nconst ZEN_KEY = \"public\";\nconst MODELS_DEV_URL = \"https://models.dev/api.json\";\nconst UPSTREAM_URL = `${UPSTREAM}/chat/completions`;\n\nconst RELAYS = [\n \"https://joddabod--2cb9984260aa11f1af5d1607ee4eb77e.web.val.run\",\n \"https://joddabod--2ce78ac260aa11f1bcfe1607ee4eb77e.web.val.run\",\n \"https://joddabod--2d4384c660aa11f195321607ee4eb77e.web.val.run\",\n \"https://joddabod--2d845b0460aa11f1b5531607ee4eb77e.web.val.run\",\n \"https://joddabod--2db5872e60aa11f1b7d71607ee4eb77e.web.val.run\",\n];\n\nlet relayIndex = 0;\n\nfunction nextRelay(): string {\n const url = RELAYS[relayIndex % RELAYS.length];\n relayIndex = (relayIndex + 1) >>> 0;\n return url;\n}\n\n// ─── CORS helpers ──────────────────────────────────────────────────────────────\n\nfunction corsHeaders(): Record {\n return {\n \"access-control-allow-origin\": \"*\",\n \"access-control-allow-methods\": \"GET, POST, OPTIONS\",\n \"access-control-allow-headers\": \"Content-Type, Authorization, x-api-key, anthropic-version\",\n \"access-control-max-age\": \"86400\",\n };\n}\n\nfunction json(data: unknown, status = 200): Response {\n return new Response(JSON.stringify(data), {\n status,\n headers: {\n \"content-type\": \"application/json\",\n ...corsHeaders(),\n },\n });\n}\n\nfunction apiError(message: string, status = 400): Response {\n return json({ error: { message, type: \"invalid_request_error\", code: null } }, status);\n}\n\n// ─── Free model list (cached) ──────────────────────────────────────────────────\n\nconst FALLBACK_MODELS = [\n { id: \"deepseek-v4-flash-free\", object: \"model\", created: 1745452800, owned_by: \"opencode-zen\", name: \"DeepSeek V4 Flash\", context_length: 200_000, capabilities: { reasoning: true, vision: false, tool_call: true }, pricing: { input: 0, output: 0 } },\n { id: \"big-pickle\", object: \"model\", created: 1729123200, owned_by: \"opencode-zen\", name: \"Big Pickle\", context_length: 200_000, capabilities: { reasoning: true, vision: false, tool_call: true }, pricing: { input: 0, output: 0 } },\n { id: \"mimo-v2.5-free\", object: \"model\", created: 1745452800, owned_by: \"opencode-zen\", name: \"MiMo 2.5\", context_length: 200_000, capabilities: { reasoning: true, vision: true, tool_call: true }, pricing: { input: 0, output: 0 } },\n];\n\nlet _modelsCache: typeof FALLBACK_MODELS | null = null;\nlet _modelsCacheAt = 0;\nconst CACHE_TTL = 5 * 60 * 1000;\n\nasync function getFreeModels() {\n const now = Date.now();\n if (_modelsCache && now - _modelsCacheAt < CACHE_TTL) return _modelsCache;\n\n try {\n const res = await fetch(MODELS_DEV_URL, { signal: AbortSignal.timeout(6000) });\n if (!res.ok) throw new Error(\"models.dev non-200\");\n const data = await res.json();\n const models = Object.values(data?.opencode?.models ?? {})\n .filter((m: any) => m.status !== \"deprecated\" && (m.cost?.input ?? 1) === 0)\n .map(toOpenAIModel);\n\n if (models.length) {\n _modelsCache = models;\n _modelsCacheAt = now;\n }\n return _modelsCache ?? FALLBACK_MODELS;\n } catch {\n return _modelsCache ?? FALLBACK_MODELS;\n }\n}\n\nfunction toOpenAIModel(m: any) {\n return {\n id: m.id,\n object: \"model\",\n created: Math.floor(new Date(m.release_date || \"2025-01-01\").getTime() / 1000),\n owned_by: \"opencode-zen\",\n name: m.name,\n context_length: m.limit?.context ?? 200_000,\n capabilities: {\n reasoning: m.reasoning ?? false,\n vision: m.attachment ?? m.modalities?.input?.some((i: string) => [\"image\", \"video\", \"audio\"].includes(i)) ?? false,\n tool_call: m.tool_call ?? true,\n },\n pricing: { input: 0, output: 0 },\n };\n}\n\n// ─── Relay fetch helper ────────────────────────────────────────────────────────\n\nasync function relayFetch(body: unknown): Promise<{ response: Response; relayUrl: string }> {\n const relayUrl = nextRelay();\n const bodyJson = JSON.stringify(body);\n const response = await fetch(relayUrl, {\n method: \"POST\",\n headers: {\n \"content-type\": \"application/json\",\n \"x-relay-target\": UPSTREAM_URL,\n authorization: `Bearer ${ZEN_KEY}`,\n },\n body: bodyJson,\n });\n return { response, relayUrl };\n}\n\nfunction copyRateLimitHeaders(upstream: Response, target: Record) {\n for (const h of [\n \"x-ratelimit-limit-requests\", \"x-ratelimit-limit-tokens\",\n \"x-ratelimit-remaining-requests\", \"x-ratelimit-remaining-tokens\",\n \"x-ratelimit-reset-requests\", \"x-ratelimit-reset-tokens\",\n ]) {\n const v = upstream.headers.get(h);\n if (v) target[h] = v;\n }\n}\n\n// ─── OpenAI endpoints ──────────────────────────────────────────────────────────\n\nasync function handleModels(): Promise {\n const models = await getFreeModels();\n return json({ object: \"list\", data: models });\n}\n\nasync function handleCompletions(req: Request): Promise {\n let body: any;\n try { body = await req.json(); } catch { return apiError(\"Request body must be valid JSON\"); }\n if (!body.model) return apiError(\"Missing required field: model\");\n if (!body.messages) return apiError(\"Missing required field: messages\");\n\n const requestedModel = body.model;\n const isStreaming = body.stream ?? false;\n\n try {\n const { response: upstream } = await relayFetch(body);\n const resHeaders: Record = { ...corsHeaders() };\n const ct = upstream.headers.get(\"content-type\");\n if (ct) resHeaders[\"content-type\"] = ct;\n copyRateLimitHeaders(upstream, resHeaders);\n\n if (!upstream.ok) {\n const errBody = await upstream.text().catch(() => \"\");\n return new Response(errBody, { status: upstream.status, headers: resHeaders });\n }\n\n if (isStreaming) {\n // For streaming, pipe through and fix the model field in each chunk\n const { readable, writable } = new TransformStream();\n const writer = writable.getWriter();\n const encoder = new TextEncoder();\n\n (async () => {\n try {\n const reader = upstream.body!.getReader();\n const decoder = new TextDecoder();\n let buf = \"\";\n\n while (true) {\n const { done, value } = await reader.read();\n if (done) break;\n buf += decoder.decode(value, { stream: true });\n const lines = buf.split(\"\\n\");\n buf = lines.pop() || \"\";\n\n for (const line of lines) {\n let outLine = line;\n const t = line.trim();\n if (t.startsWith(\"data: \") && t !== \"data: [DONE]\") {\n try {\n const chunk = JSON.parse(t.slice(6));\n if (chunk.model) chunk.model = requestedModel;\n if (chunk.choices?.[0]?.delta) {\n // Some streaming chunks include model in the delta, patch it\n }\n outLine = `data: ${JSON.stringify(chunk)}`;\n } catch { /* pass through */ }\n }\n await writer.write(encoder.encode(outLine + \"\\n\"));\n }\n }\n } catch { /* stream interrupted */ }\n await writer.close();\n })();\n\n return new Response(readable, { status: upstream.status, headers: resHeaders });\n } else {\n // Non-streaming: read body, fix model name, return\n const upstreamData = await upstream.json().catch(() => null);\n if (!upstreamData) {\n return new Response(await upstream.text().catch(() => \"\"), { status: upstream.status, headers: resHeaders });\n }\n if (upstreamData.model) upstreamData.model = requestedModel;\n return json(upstreamData, upstream.status);\n }\n } catch (err: any) {\n return apiError(`Relay failed: ${err.message}`, 502);\n }\n}\n\n// ─── Anthropic endpoint ────────────────────────────────────────────────────────\n\n/** Extract auth token from either OpenAI-style (Authorization) or Anthropic-style (x-api-key) */\nfunction extractBearer(req: Request): string {\n const auth = req.headers.get(\"authorization\") || \"\";\n if (auth.startsWith(\"Bearer \")) return auth.slice(7);\n const key = req.headers.get(\"x-api-key\") || \"\";\n return key;\n}\n\n/** Flatten Anthropic content blocks into a single text string */\nfunction flattenAnthropicContent(content: any): string {\n if (typeof content === \"string\") return content;\n if (Array.isArray(content)) {\n return content\n .filter((b: any) => b.type === \"text\" || b.type === \"image_url\" || b.type === \"image\")\n .map((b: any) => {\n if (b.type === \"text\") return b.text;\n return \"[image]\";\n })\n .join(\"\\n\");\n }\n return String(content);\n}\n\n/** Convert an Anthropic /v1/messages request body to OpenAI /chat/completions format */\nfunction anthropicToOpenAI(body: any): any {\n const openAIMessages: any[] = [];\n\n // System prompt → system message\n if (body.system) {\n if (typeof body.system === \"string\") {\n openAIMessages.push({ role: \"system\", content: body.system });\n } else if (Array.isArray(body.system)) {\n const text = body.system.filter((b: any) => b.type === \"text\").map((b: any) => b.text).join(\"\\n\");\n if (text) openAIMessages.push({ role: \"system\", content: text });\n }\n }\n\n // Messages\n for (const msg of body.messages || []) {\n const content = flattenAnthropicContent(msg.content);\n if (content) {\n openAIMessages.push({ role: msg.role, content });\n }\n }\n\n return {\n model: body.model,\n messages: openAIMessages,\n max_tokens: body.max_tokens ?? 4096,\n temperature: body.temperature,\n top_p: body.top_p,\n stream: body.stream ?? false,\n };\n}\n\n/** Mapping from OpenAI finish_reason to Anthropic stop_reason */\nfunction mapStopReason(finishReason: string | null): string {\n switch (finishReason) {\n case \"stop\": return \"end_turn\";\n case \"length\": return \"max_tokens\";\n case \"content_filter\": return \"content_filtered\";\n case \"tool_calls\": return \"tool_use\";\n default: return \"end_turn\";\n }\n}\n\n/** Convert an OpenAI non-streaming response to Anthropic /v1/messages format */\nfunction openAIToAnthropic(openAIRes: any, model: string): any {\n const choice = openAIRes.choices?.[0] || {};\n const message = choice.message || {};\n const usage = openAIRes.usage || {};\n\n return {\n id: openAIRes.id || `msg_${crypto.randomUUID()}`,\n type: \"message\",\n role: \"assistant\",\n content: [\n {\n type: \"text\",\n text: message.content || \"\",\n },\n ],\n model: model,\n stop_reason: mapStopReason(choice.finish_reason),\n stop_sequence: null,\n usage: {\n input_tokens: usage.prompt_tokens || 0,\n output_tokens: usage.completion_tokens || 0,\n },\n };\n}\n\n/** Translate one OpenAI SSE chunk into zero or more Anthropic SSE strings */\nfunction translateStreamChunk(\n openAIChunk: any,\n model: string,\n state: { messageStarted: boolean; contentStarted: boolean; contentIndex: number },\n): string[] {\n const events: string[] = [];\n const delta = openAIChunk.choices?.[0]?.delta || {};\n const finishReason = openAIChunk.choices?.[0]?.finish_reason || null;\n const usage = openAIChunk.usage || null;\n\n // message_start\n if (!state.messageStarted) {\n state.messageStarted = true;\n events.push(`event: message_start\ndata: ${JSON.stringify({\n type: \"message_start\",\n message: {\n id: `msg_${crypto.randomUUID()}`,\n type: \"message\",\n role: \"assistant\",\n content: [],\n model,\n stop_reason: null,\n stop_sequence: null,\n usage: { input_tokens: 0, output_tokens: 0 },\n },\n})}\n\n`);\n }\n\n // content_block_start\n if (delta.content && !state.contentStarted) {\n state.contentStarted = true;\n events.push(`event: content_block_start\ndata: ${JSON.stringify({\n type: \"content_block_start\",\n index: state.contentIndex,\n content_block: { type: \"text\", text: \"\" },\n})}\n\n`);\n }\n\n // content_block_delta\n if (delta.content) {\n events.push(`event: content_block_delta\ndata: ${JSON.stringify({\n type: \"content_block_delta\",\n index: state.contentIndex,\n delta: { type: \"text_delta\", text: delta.content },\n})}\n\n`);\n }\n\n // reasoning\n if (delta.reasoning_content) {\n events.push(`event: content_block_delta\ndata: ${JSON.stringify({\n type: \"content_block_delta\",\n index: state.contentIndex,\n delta: { type: \"reasoning_delta\", reasoning: delta.reasoning_content },\n})}\n\n`);\n }\n\n // content_block_stop + message_delta on finish\n if (finishReason && state.contentStarted) {\n events.push(`event: content_block_stop\ndata: ${JSON.stringify({ type: \"content_block_stop\", index: state.contentIndex })}\n\n`);\n events.push(`event: message_delta\ndata: ${JSON.stringify({\n type: \"message_delta\",\n delta: { stop_reason: mapStopReason(finishReason), stop_sequence: null },\n usage: usage ? { output_tokens: usage.completion_tokens || 0 } : undefined,\n})}\n\n`);\n }\n\n // message_stop\n if (finishReason) {\n events.push(`event: message_stop\ndata: ${JSON.stringify({ type: \"message_stop\" })}\n\n`);\n }\n\n return events;\n}\n\nasync function handleAnthropicMessages(req: Request): Promise {\n // Read auth token\n const bearer = extractBearer(req);\n\n let body: any;\n try { body = await req.json(); } catch { return apiError(\"Request body must be valid JSON\"); }\n if (!body.model) return apiError(\"Missing required field: model\");\n if (!body.messages) return apiError(\"Missing required field: messages\");\n if (!body.max_tokens) return apiError(\"Missing required field: max_tokens\");\n\n const model = body.model;\n const streaming = body.stream ?? false;\n const openAIBody = anthropicToOpenAI(body);\n\n if (streaming) {\n // Streaming: read the OpenAI stream and translate each chunk\n try {\n const relayUrl = nextRelay();\n const upstream = await fetch(relayUrl, {\n method: \"POST\",\n headers: {\n \"content-type\": \"application/json\",\n \"x-relay-target\": UPSTREAM_URL,\n authorization: `Bearer ${ZEN_KEY}`,\n },\n body: JSON.stringify(openAIBody),\n });\n\n if (!upstream.ok) {\n const errText = await upstream.text().catch(() => \"Upstream error\");\n return json(\n { error: { type: \"api_error\", message: errText.slice(0, 200) } },\n upstream.status,\n );\n }\n\n const { readable, writable } = new TransformStream();\n const writer = writable.getWriter();\n const encoder = new TextEncoder();\n\n const state = { messageStarted: false, contentStarted: false, contentIndex: 0 };\n\n (async () => {\n try {\n const reader = upstream.body!.getReader();\n const decoder = new TextDecoder();\n let buf = \"\";\n\n while (true) {\n const { done, value } = await reader.read();\n if (done) break;\n buf += decoder.decode(value, { stream: true });\n const lines = buf.split(\"\\n\");\n buf = lines.pop() || \"\";\n\n for (const line of lines) {\n const t = line.trim();\n if (!t || t === \"data: [DONE]\" || !t.startsWith(\"data: \")) continue;\n try {\n const chunk = JSON.parse(t.slice(6));\n const events = translateStreamChunk(chunk, model, state);\n for (const evt of events) {\n await writer.write(encoder.encode(evt));\n }\n } catch { /* skip malformed */ }\n }\n }\n // If stream ended but we never got a finish_reason, close cleanly\n if (state.contentStarted) {\n await writer.write(encoder.encode(`event: content_block_stop\\ndata: ${JSON.stringify({ type: \"content_block_stop\", index: 0 })}\\n\\n`));\n await writer.write(encoder.encode(`event: message_delta\\ndata: ${JSON.stringify({ type: \"message_delta\", delta: { stop_reason: \"end_turn\", stop_sequence: null } })}\\n\\n`));\n await writer.write(encoder.encode(`event: message_stop\\ndata: ${JSON.stringify({ type: \"message_stop\" })}\\n\\n`));\n }\n } catch { /* stream interrupted */ }\n await writer.close();\n })();\n\n return new Response(readable, {\n headers: {\n \"content-type\": \"text/event-stream\",\n \"cache-control\": \"no-cache\",\n ...corsHeaders(),\n },\n });\n } catch (err: any) {\n return apiError(`Relay failed: ${err.message}`, 502);\n }\n } else {\n // Non-streaming: proxy and convert response\n try {\n const { response: upstream } = await relayFetch(openAIBody);\n const upstreamData = await upstream.json().catch(() => null);\n\n if (!upstream.ok || !upstreamData) {\n const msg = upstreamData?.error?.message || `Upstream error ${upstream.status}`;\n return json(\n { error: { type: \"api_error\", message: msg } },\n upstream.status,\n );\n }\n\n const anthropicRes = openAIToAnthropic(upstreamData, model);\n const resHeaders: Record = { ...corsHeaders() };\n copyRateLimitHeaders(upstream, resHeaders);\n return json(anthropicRes, upstream.status);\n } catch (err: any) {\n return apiError(`Relay failed: ${err.message}`, 502);\n }\n }\n}\n\n// ─── Router ─────────────────────────────────────────────────────────────────────\n\nexport default async function handler(req: Request): Promise {\n const url = new URL(req.url);\n const method = req.method;\n\n if (method === \"OPTIONS\") {\n return new Response(null, { status: 204, headers: corsHeaders() });\n }\n\n if (method === \"GET\" && url.pathname === \"/v1/models\") {\n return handleModels();\n }\n\n if (method === \"POST\" && url.pathname === \"/v1/chat/completions\") {\n return handleCompletions(req);\n }\n\n // Anthropic-compatible endpoint\n if (method === \"POST\" && url.pathname === \"/v1/messages\") {\n return handleAnthropicMessages(req);\n }\n\n return apiError(\"Not found\", 404);\n}\n","createdAt",["D",1780646078930],"branchVersion","branchId","ab3c45e1-4cb9-411c-9399-278023ecc3f7","id","d5ff3af5-511d-43c2-8afd-3166f8ba72d4","relays",["D",1780641307836],"deletedAt","legacyEnvVars","privacy","public","httpPrivacy","description","imageUrl","c9497c68-4a62-4420-ba65-1111208a07d6","pinnedFileId","pinPreview","pinMigrated","screenshotVersion","likeCount",0,"orgId","019e967b-e157-7346-86bd-5e83fe7256ac","tursoName","tursoHostname","tursoMeasuredAt","pinned","userPerm","r","valPerm","blobPerm","rw","emailPerm","sqlitePerm","projectPerm","telemetryPerm","user",{"_94":93,"_102":98},"org",{"_94":93,"_95":96,"_97":98,"_18":54,"_99":100,"_101":25},"branch",{"_13":83,"_32":84,"_43":42,"_85":44,"_86":-5,"_87":-5,"_88":-5,"_89":-5,"_90":-5,"_91":45,"_92":93},"isOwned",17,"main","projectId","forkedBranchId","forkedBranchVersion","forkedProjectPrivacy","forkedProjectId","forkedProjectOrgId","projectName","userHandle","joddabod","handle","displayName","Joud Aljawish","avatar","https://lh3.googleusercontent.com/a/ACg8ocIxU-BTO7FNgB5nFxdnvwMbtFz65DJETXIEYKpXDhUFbhRnk6FH=s96-c","tier","free","canRestrictVals","profileImageUrl","branchName","","adoptOrgId",{"_43":44,"_32":45,"_38":150,"_47":-5,"_48":25,"_49":50,"_51":50,"_52":-5,"_53":-5,"_18":54,"_55":-5,"_56":25,"_57":-5,"_58":-5,"_59":60,"_61":62,"_63":-5,"_64":-5,"_65":-5,"_66":-5,"_67":68,"_69":68,"_70":71,"_72":71,"_73":71,"_74":68,"_75":68,"_76":151,"_78":152,"_80":153,"_82":25},"tree",[145],"files",[145],"counts",{"_134":60,"_135":136,"_137":60,"_138":139,"_140":141},"contents",[126],"upstreamProject","upstreamBranch","emptyEnvVars","flags",[],"hasHttpFiles",true,"grants","orgs","customDomains",{},{"_43":127,"_32":28,"_30":31,"_128":-5,"_41":42,"_40":14,"_129":25,"_38":130,"_13":131,"_33":34,"_35":37,"_24":25,"_132":-5,"_27":133},"49bb72cf-2aae-443e-9387-63d86b2373a7","parentId","latest",["D",1780646078930],5,"label",[28],"pulls","branches",1,"forks","envVars",{"_142":60,"_144":60},"connections",{"_142":60,"_143":60},"count","unfilledCount","emptyCount",{"_43":127,"_32":28,"_30":31,"_128":-5,"_41":42,"_40":14,"_129":25,"_38":146,"_13":131,"_33":34,"_35":37,"_24":25,"_132":-5,"_147":148,"_27":149},["D",1780646078930],"children",[],[28],["D",1780641307836],{"_94":93,"_102":98},{"_94":93,"_95":96,"_97":98,"_18":54,"_99":100,"_101":25},{"_13":83,"_32":84,"_43":42,"_85":44,"_86":-5,"_87":-5,"_88":-5,"_89":-5,"_90":-5,"_91":45,"_92":93},"ctx",{"_76":-5,"_123":187,"_188":189,"_190":191,"_192":25,"_193":25,"_194":25,"_195":121,"_196":-5,"_197":-7,"_198":-7,"_199":60,"_200":25,"_201":25,"_202":-5,"_203":25,"_204":25,"_205":25,"_206":-5},"ENV",{"_158":159,"_160":161,"_162":84,"_163":164,"_165":166,"_167":168,"_169":170,"_171":172,"_173":174,"_175":176,"_177":178,"_179":180,"_181":182,"_183":184,"_185":186},"REACT_APP_SENTRY_DSN","https://586020d86f05489db63a554690acf1e9@o1398359.ingest.sentry.io/6724499","RENDER_GIT_COMMIT","e87eb04e707cf64e7e532c8e8ec395f209e17f15","RENDER_GIT_BRANCH","POSTHOG_API_TOKEN","phc_nmoyuAkStEZuUp69rYlg1DZgrIt19Gx0Y4Hwf8tGgJP","POSTHOG_API_HOST","https://wart.val.town","WEB_API_TEMPLATE","https://{val}.web.val.run","APP_URL","https://www.val.town","MODULE_URL","https://esm.town","VANITY_WEB_API_TEMPLATE","https://{label}.val.run/","DENOLS_URL","https://lsp.val.town","NODE_ENV","production","CLOUDFLARE_TURNSTILE_SITE_KEY","0x4AAAAAACtaIZJH7s9rBMZ6","GOOGLE_CLIENT_ID","875989696899-jtsqkhlbdv5jrlsk8631p7ipjg3c6o62.apps.googleusercontent.com","STRIPE_PUBLISHABLE_KEY","pk_live_51N2zvWEpZTTLipIkVEwXrUG9AT0yeOOEYxszarZT1jpsZV58Ly6wlg3SltZpHN26CUalOBmUSY7adQRQiSkCZz2500qSXNkf8Y","API_URL","https://api.val.town",[],"preferencesKeybindings","default","preferencesTheme","light","preferencesFixers","preferencesCopilot","preferencesSelectionAi","preferencesCloseBrackets","preferencesChatPanelState","preferencesDprint","preferencesDenofmt","notificationCount","flagDenols","isBusinessOrgMember","businessOrgHandle","isBusinessOrgAdmin","isHigherTownieTier","navCollapsed","currentOrgId"]

Diff · v16