`, favicon, and body of `Root()` to\nmatch your app's existing `index.html`, and stamp anything else your shell\nreferences directly (images, css) with `immutableFileUrl()`.\n\n1. Create `frontend/root.tsx`:\n\n```tsx\n/** @jsxImportSource npm:hono@4/jsx */\nimport { raw } from \"npm:hono@4/html\";\nimport { immutableFileUrl } from \"https://esm.town/v/std/utils/index.ts\";\n\n// The HTML shell for the app. It is never cached, so immutableFileUrl can stamp\n// the current val version into each asset URL — the browser caches those\n// immutably, and publishing the val changes the URLs, refetching exactly once.\nexport function Root() {\n return (\n <>\n {raw(\"<!DOCTYPE html>\")}\n <html lang=\"en\">\n <head>\n <meta charSet=\"UTF-8\" />\n <meta name=\"viewport\" content=\"width=device-width, initial-scale=1.0\" />\n <title>React Hono Val Town Starter

[{"_1":2,"_3":-5,"_4":-5},"loaderData",{"_5":6,"_7":8,"_9":10},"actionData","errors","root",{"_340":341,"_342":343},"routes/_app.x.$handle.$name",{"_105":104,"_38":51,"_113":95,"_33":107,"_13":14,"_114":-5,"_11":115,"_116":117,"_118":119,"_120":121,"_122":123,"_26":-5,"_124":-5,"_125":-5,"_126":31,"_127":128,"_129":64,"_130":-5,"_131":-5,"_132":133},"routes/_app.x.$handle.$name.branch.$branch.version.$version.diff",{"_11":12,"_13":14,"_15":16,"_17":18,"_19":-5},"project",{"_49":50,"_38":51,"_44":52,"_53":-5,"_54":31,"_55":56,"_57":56,"_58":59,"_60":-5,"_19":61,"_62":37,"_63":64,"_65":-5,"_66":-5,"_67":68,"_69":70,"_71":72,"_73":74,"_75":76,"_77":-5,"_78":79,"_80":79,"_81":82,"_83":82,"_84":82,"_85":79,"_86":79,"_87":88,"_89":90,"_91":92,"_93":31},"version",16,"changes",[25],"aiChatInfo",{"_20":21,"_22":-5,"_23":-5,"_24":-5},"userId","aiChatMessageId","bd0ef1ff-c864-4254-a616-a631f038a507","aiChatThreadId","aiChatThreadUserId","aiChatThreadTitle",{"_26":27,"_28":29,"_30":31,"_32":31,"_33":34,"_35":34},"file",{"_36":37,"_38":34,"_39":40,"_41":43,"_44":45,"_46":14,"_47":48,"_30":31},"previous",{"_36":37,"_38":34,"_39":40,"_41":42},"deleted",false,"deletedHere","path","index.ts","latestPath","fileId","01a0aada-f650-726c-8e46-8ec8db0c6ebd","name","fileType","http","content","import {\n parseVal,\n serveImmutableFile,\n} from \"https://esm.town/v/std/utils/index.ts\";\nimport { Hono } from \"npm:hono\";\nimport { Root } from \"./frontend/root.tsx\";\nimport { findModel, MODELS } from \"./frontend/models.ts\";\nimport {\n extractMediaUrl,\n falConfigured,\n getResult,\n getStatus,\n submit,\n} from \"./fal.ts\";\nimport {\n deleteGeneration,\n getGeneration,\n insertGeneration,\n listGenerations,\n markDone,\n markFailed,\n markRunning,\n} from \"./db.ts\";\n\nconst app = new Hono();\n\n// ---------------------------------------------------------------- frontend\n\n// Serve the (never-cached) HTML shell at the root /\napp.get(\"/\", (c) => c.html(Root()));\n\n// Serve frontend files at the version-addressed URLs Root() stamps, cached immutably\napp.get(\"/__immutable/*\", (c) => serveImmutableFile(c.req.path));\n\n// View source redirect\napp.get(\"/source\", (c) => c.redirect(parseVal().links.self.val));\n\n// ---------------------------------------------------------------- api\n\n/** Is the val usable yet? The UI shows setup instructions when this is false. */\napp.get(\"/api/config\", (c) => c.json({ falConfigured: falConfigured() }));\n\napp.get(\"/api/models\", (c) => c.json({ models: MODELS }));\n\napp.get(\"/api/generations\", async (c) => {\n const generations = await listGenerations();\n return c.json({ generations });\n});\n\n/**\n * Submit a new generation. Returns as soon as fal accepts the job — the row\n * lands in the DB as 'queued' and the client polls /refresh until it settles.\n */\napp.post(\"/api/generate\", async (c) => {\n if (!falConfigured()) {\n return c.json(\n { error: \"FAL_KEY is not set on this val. Add it, then reload the page.\" },\n 503,\n );\n }\n\n let body: { modelId?: unknown; prompt?: unknown; extra?: unknown };\n try {\n body = await c.req.json();\n } catch {\n return c.json({ error: \"Request body must be JSON\" }, 400);\n }\n\n const modelId = typeof body.modelId === \"string\" ? body.modelId : \"\";\n const model = findModel(modelId);\n if (!model) {\n return c.json({ error: `Unknown model: ${modelId || \"(missing)\"}` }, 400);\n }\n\n const prompt = typeof body.prompt === \"string\" ? body.prompt.trim() : \"\";\n if (!prompt) {\n return c.json({ error: \"Prompt is required\" }, 400);\n }\n if (prompt.length > 4000) {\n return c.json({ error: \"Prompt is too long (4000 char max)\" }, 400);\n }\n\n // Escape hatch for model-specific knobs we don't surface in the UI.\n let extra: Record = {};\n if (body.extra && typeof body.extra === \"object\") {\n extra = body.extra as Record;\n }\n\n const input: Record = {\n ...model.defaults,\n ...extra,\n [model.promptField]: prompt,\n };\n\n try {\n const submitted = await submit(model.id, input);\n const generation = await insertGeneration({\n modality: model.modality,\n model: model.id,\n modelLabel: model.label,\n prompt,\n requestId: submitted.request_id,\n statusUrl: submitted.status_url,\n responseUrl: submitted.response_url,\n });\n return c.json({ generation }, 201);\n } catch (err) {\n return c.json({ error: errorMessage(err) }, 502);\n }\n});\n\n/**\n * Poll one job. The client calls this on a timer for anything not yet settled.\n * Once fal reports COMPLETED we fetch the payload, pull the media URL out, and\n * persist it — so a reload never re-hits fal for a finished job.\n */\napp.post(\"/api/generations/:id/refresh\", async (c) => {\n const id = Number(c.req.param(\"id\"));\n if (!Number.isInteger(id)) return c.json({ error: \"Bad id\" }, 400);\n\n const generation = await getGeneration(id);\n if (!generation) return c.json({ error: \"Not found\" }, 404);\n\n // Terminal states are already final — just hand the row back.\n if (generation.status === \"done\" || generation.status === \"failed\") {\n return c.json({ generation });\n }\n if (!generation.statusUrl || !generation.responseUrl) {\n await markFailed(id, \"Missing fal status URL — resubmit this job.\");\n return c.json({ generation: await getGeneration(id) });\n }\n\n try {\n const status = await getStatus(generation.statusUrl);\n\n if (status.status === \"IN_QUEUE\") {\n return c.json({ generation });\n }\n\n if (status.status === \"IN_PROGRESS\") {\n await markRunning(id);\n return c.json({ generation: await getGeneration(id) });\n }\n\n // COMPLETED — but \"completed\" can still carry a model-level error.\n if (status.error) {\n await markFailed(id, status.error);\n return c.json({ generation: await getGeneration(id) });\n }\n\n const model = findModel(generation.model);\n const payload = await getResult(generation.responseUrl);\n const url = extractMediaUrl(\n payload,\n model?.resultPath ?? \"\",\n (model?.modality ?? generation.modality) as \"image\" | \"video\" | \"audio\",\n );\n\n if (!url) {\n await markFailed(\n id,\n \"fal returned no media URL. The model may have changed its response shape.\",\n );\n } else {\n await markDone(id, url);\n }\n return c.json({ generation: await getGeneration(id) });\n } catch (err) {\n await markFailed(id, errorMessage(err));\n return c.json({ generation: await getGeneration(id) });\n }\n});\n\napp.delete(\"/api/generations/:id\", async (c) => {\n const id = Number(c.req.param(\"id\"));\n if (!Number.isInteger(id)) return c.json({ error: \"Bad id\" }, 400);\n await deleteGeneration(id);\n return c.json({ ok: true });\n});\n\n// ---------------------------------------------------------------- plumbing\n\n// Unwrap and rethrow Hono errors as the original error\n// If you delete this, all errors will be \"Internal Server Error\"\napp.onError((err) => Promise.reject(err));\n\nfunction errorMessage(err: unknown): string {\n return err instanceof Error ? err.message : String(err);\n}\n\n// HTTP vals expect an exported \"fetch handler\"\n// This is how you \"run the server\" in Val Town with Hono\nexport default app.fetch;","import {\n parseVal,\n serveImmutableFile,\n} from \"https://esm.town/v/std/utils/index.ts\";\nimport { Hono } from \"npm:hono\";\nimport { Root } from \"./frontend/root.tsx\";\nimport { findModel, MODELS } from \"./frontend/models.ts\";\nimport {\n extractMediaUrl,\n falConfigured,\n getResult,\n getStatus,\n submit,\n} from \"./fal.ts\";\nimport {\n deleteGeneration,\n getGeneration,\n insertGeneration,\n listGenerations,\n markDone,\n markFailed,\n markRunning,\n} from \"./db.ts\";\n\nconst app = new Hono();\n\n// ---------------------------------------------------------------- frontend\n\n// Serve the (never-cached) HTML shell at the root /\napp.get(\"/\", (c) => c.html(Root()));\n\n// Serve frontend files at the version-addressed URLs Root() stamps, cached immutably\napp.get(\"/__immutable/*\", (c) => serveImmutableFile(c.req.path));\n\n// View source redirect\napp.get(\"/source\", (c) => c.redirect(parseVal().links.self.val));\n\n// ---------------------------------------------------------------- api\n\n/**\n * Is the val usable yet? The UI shows setup instructions when this is false.\n * envVarUrl is derived from the val's own URL so a remix points at the remix's\n * editor rather than ours.\n */\napp.get(\"/api/config\", (c) =>\n c.json({\n falConfigured: falConfigured(),\n envVarUrl: `${parseVal().links.self.val}/environment-variables?key=FAL_KEY`,\n }));\n\napp.get(\"/api/models\", (c) => c.json({ models: MODELS }));\n\napp.get(\"/api/generations\", async (c) => {\n const generations = await listGenerations();\n return c.json({ generations });\n});\n\n/**\n * Submit a new generation. Returns as soon as fal accepts the job — the row\n * lands in the DB as 'queued' and the client polls /refresh until it settles.\n */\napp.post(\"/api/generate\", async (c) => {\n if (!falConfigured()) {\n return c.json(\n { error: \"FAL_KEY is not set on this val. Add it, then reload the page.\" },\n 503,\n );\n }\n\n let body: { modelId?: unknown; prompt?: unknown; extra?: unknown };\n try {\n body = await c.req.json();\n } catch {\n return c.json({ error: \"Request body must be JSON\" }, 400);\n }\n\n const modelId = typeof body.modelId === \"string\" ? body.modelId : \"\";\n const model = findModel(modelId);\n if (!model) {\n return c.json({ error: `Unknown model: ${modelId || \"(missing)\"}` }, 400);\n }\n\n const prompt = typeof body.prompt === \"string\" ? body.prompt.trim() : \"\";\n if (!prompt) {\n return c.json({ error: \"Prompt is required\" }, 400);\n }\n if (prompt.length > 4000) {\n return c.json({ error: \"Prompt is too long (4000 char max)\" }, 400);\n }\n\n // Escape hatch for model-specific knobs we don't surface in the UI.\n let extra: Record = {};\n if (body.extra && typeof body.extra === \"object\") {\n extra = body.extra as Record;\n }\n\n const input: Record = {\n ...model.defaults,\n ...extra,\n [model.promptField]: prompt,\n };\n\n try {\n const submitted = await submit(model.id, input);\n const generation = await insertGeneration({\n modality: model.modality,\n model: model.id,\n modelLabel: model.label,\n prompt,\n requestId: submitted.request_id,\n statusUrl: submitted.status_url,\n responseUrl: submitted.response_url,\n });\n return c.json({ generation }, 201);\n } catch (err) {\n return c.json({ error: errorMessage(err) }, 502);\n }\n});\n\n/**\n * Poll one job. The client calls this on a timer for anything not yet settled.\n * Once fal reports COMPLETED we fetch the payload, pull the media URL out, and\n * persist it — so a reload never re-hits fal for a finished job.\n */\napp.post(\"/api/generations/:id/refresh\", async (c) => {\n const id = Number(c.req.param(\"id\"));\n if (!Number.isInteger(id)) return c.json({ error: \"Bad id\" }, 400);\n\n const generation = await getGeneration(id);\n if (!generation) return c.json({ error: \"Not found\" }, 404);\n\n // Terminal states are already final — just hand the row back.\n if (generation.status === \"done\" || generation.status === \"failed\") {\n return c.json({ generation });\n }\n if (!generation.statusUrl || !generation.responseUrl) {\n await markFailed(id, \"Missing fal status URL — resubmit this job.\");\n return c.json({ generation: await getGeneration(id) });\n }\n\n try {\n const status = await getStatus(generation.statusUrl);\n\n if (status.status === \"IN_QUEUE\") {\n return c.json({ generation });\n }\n\n if (status.status === \"IN_PROGRESS\") {\n await markRunning(id);\n return c.json({ generation: await getGeneration(id) });\n }\n\n // COMPLETED — but \"completed\" can still carry a model-level error.\n if (status.error) {\n await markFailed(id, status.error);\n return c.json({ generation: await getGeneration(id) });\n }\n\n const model = findModel(generation.model);\n const payload = await getResult(generation.responseUrl);\n const url = extractMediaUrl(\n payload,\n model?.resultPath ?? \"\",\n (model?.modality ?? generation.modality) as \"image\" | \"video\" | \"audio\",\n );\n\n if (!url) {\n await markFailed(\n id,\n \"fal returned no media URL. The model may have changed its response shape.\",\n );\n } else {\n await markDone(id, url);\n }\n return c.json({ generation: await getGeneration(id) });\n } catch (err) {\n await markFailed(id, errorMessage(err));\n return c.json({ generation: await getGeneration(id) });\n }\n});\n\napp.delete(\"/api/generations/:id\", async (c) => {\n const id = Number(c.req.param(\"id\"));\n if (!Number.isInteger(id)) return c.json({ error: \"Bad id\" }, 400);\n await deleteGeneration(id);\n return c.json({ ok: true });\n});\n\n// ---------------------------------------------------------------- plumbing\n\n// Unwrap and rethrow Hono errors as the original error\n// If you delete this, all errors will be \"Internal Server Error\"\napp.onError((err) => Promise.reject(err));\n\nfunction errorMessage(err: unknown): string {\n return err instanceof Error ? err.message : String(err);\n}\n\n// HTTP vals expect an exported \"fetch handler\"\n// This is how you \"run the server\" in Val Town with Hono\nexport default app.fetch;","createdAt",["D",1789573055015],"branchVersion","branchId","5659524a-415a-4634-89d5-e97fabee9dad","id","9d13aa12-d771-4e42-a8aa-33676d4852dc","ai-creative-studio",["D",1789572871721],"deletedAt","legacyEnvVars","privacy","public","httpPrivacy","description","Generate images, video and audio from a prompt","imageUrl","c6f5037f-1586-4b70-8412-cc86b891d03e","pinnedFileId","pinPreview",true,"pinMigrated","screenshotVersion","likeCount",0,"orgId","01a0aa85-bc29-728f-b440-3fa04bc3896e","tursoName","9d13aa12d7714e42a8aa33676d4852dc","tursoHostname","9d13aa12d7714e42a8aa33676d4852dc-val-town.aws-us-east-2.turso.io","tursoMeasuredAt",["D",1789573271003],"pinned","userPerm","r","valPerm","blobPerm","rw","emailPerm","sqlitePerm","projectPerm","telemetryPerm","user",{"_105":104,"_112":107},"org",{"_105":104,"_106":107,"_108":107,"_19":61,"_109":110,"_111":31},"branch",{"_13":94,"_38":95,"_49":48,"_96":50,"_97":-5,"_98":-5,"_99":-5,"_100":-5,"_101":-5,"_102":51,"_103":104},"isOwned",18,"main","projectId","forkedBranchId","forkedBranchVersion","forkedProjectPrivacy","forkedProjectId","forkedProjectOrgId","projectName","userHandle","mansurkurtaran51","handle","displayName","","avatar","tier","free","canRestrictVals","profileImageUrl","branchName","adoptOrgId",{"_49":50,"_38":51,"_44":335,"_53":-5,"_54":31,"_55":56,"_57":56,"_58":59,"_60":-5,"_19":61,"_62":37,"_63":64,"_65":-5,"_66":-5,"_67":68,"_69":70,"_71":72,"_73":74,"_75":336,"_77":-5,"_78":79,"_80":79,"_81":82,"_83":82,"_84":82,"_85":79,"_86":79,"_87":337,"_89":338,"_91":339,"_93":31},"tree",[191,206,207,208,209,210],"files",[191,192,193,194,195,196,197,198,199,200,201,202,203,204,205,206,207,208,209,210],"counts",{"_181":68,"_182":160,"_183":68,"_184":185,"_186":187},"contents",[134,135,136,137,138,139],"upstreamProject","upstreamBranch","emptyEnvVars","flags",[],"hasHttpFiles","grants","orgs","customDomains",{},{"_49":175,"_38":176,"_36":177,"_141":-5,"_47":48,"_46":160,"_142":64,"_44":178,"_13":68,"_39":179,"_41":-5,"_30":31,"_145":-5,"_33":180},{"_49":169,"_38":170,"_36":171,"_141":-5,"_47":48,"_46":160,"_142":64,"_44":172,"_13":68,"_39":26,"_41":173,"_30":31,"_145":-5,"_33":174},{"_49":163,"_38":164,"_36":165,"_141":-5,"_47":48,"_46":160,"_142":64,"_44":166,"_13":68,"_39":26,"_41":167,"_30":31,"_145":-5,"_33":168},{"_49":155,"_38":156,"_36":157,"_141":-5,"_47":48,"_46":158,"_142":64,"_44":159,"_13":160,"_39":152,"_41":161,"_30":31,"_145":-5,"_33":162},{"_49":147,"_38":148,"_36":149,"_141":-5,"_47":48,"_46":150,"_142":64,"_44":151,"_13":68,"_39":152,"_41":153,"_30":31,"_145":-5,"_33":154},{"_49":140,"_38":34,"_36":37,"_141":-5,"_47":48,"_46":14,"_142":64,"_44":143,"_13":144,"_39":40,"_41":43,"_30":31,"_145":-5,"_33":146},"804f91a4-da6a-4ffc-bdcf-22ff5fccc98f","parentId","latest",["D",1789573055015],2,"label",[34],"d916a802-b42d-4eca-b71d-1478834777d8","fal.ts","45789faa-b1e4-11f1-b1f3-1607ee4eb77e",3,["D",1789572940131],"script","// Thin REST client for fal.ai's queue API. No SDK — just fetch, so there's\n// nothing to keep in sync with npm releases.\n//\n// Lifecycle: submit() → getStatus() (poll) → getResult()\n// Docs: https://fal.ai/docs/documentation/model-apis/inference/queue\n\nconst QUEUE_BASE = \"https://queue.fal.run\";\n\n/** fal keys are per-account. Read at call time so a missing key is a clear error. */\nexport function falKey(): string | undefined {\n return Deno.env.get(\"FAL_KEY\") ?? undefined;\n}\n\nexport function falConfigured(): boolean {\n return Boolean(falKey());\n}\n\nfunction authHeaders(): HeadersInit {\n const key = falKey();\n if (!key) {\n throw new Error(\n \"FAL_KEY is not set on this val. Add it, then reload the page.\",\n );\n }\n return {\n Authorization: `Key ${key}`,\n \"Content-Type\": \"application/json\",\n };\n}\n\nexport type SubmitResult = {\n request_id: string;\n status_url: string;\n response_url: string;\n cancel_url?: string;\n queue_position?: number;\n};\n\nexport type QueueStatus = {\n status: \"IN_QUEUE\" | \"IN_PROGRESS\" | \"COMPLETED\";\n queue_position?: number;\n logs?: { message: string }[];\n metrics?: { inference_time?: number };\n error?: string;\n error_type?: string;\n};\n\n/** Send a request to the queue. Returns immediately — the work happens async. */\nexport async function submit(\n modelId: string,\n input: Record,\n): Promise {\n const res = await fetch(`${QUEUE_BASE}/${modelId}`, {\n method: \"POST\",\n headers: authHeaders(),\n body: JSON.stringify(input),\n });\n\n const text = await res.text();\n if (!res.ok) {\n throw new Error(`fal submit failed (${res.status}): ${trim(text)}`);\n }\n\n try {\n return JSON.parse(text) as SubmitResult;\n } catch {\n throw new Error(`fal submit returned non-JSON: ${trim(text)}`);\n }\n}\n\n/** Poll the status URL fal handed back at submit time. */\nexport async function getStatus(statusUrl: string): Promise {\n const res = await fetch(`${statusUrl}?logs=1`, { headers: authHeaders() });\n const text = await res.text();\n\n if (!res.ok) {\n throw new Error(`fal status failed (${res.status}): ${trim(text)}`);\n }\n return JSON.parse(text) as QueueStatus;\n}\n\n/** Fetch the finished payload. Only meaningful once status is COMPLETED. */\nexport async function getResult(\n responseUrl: string,\n): Promise> {\n const res = await fetch(responseUrl, { headers: authHeaders() });\n const text = await res.text();\n\n if (!res.ok) {\n throw new Error(`fal result failed (${res.status}): ${trim(text)}`);\n }\n return JSON.parse(text) as Record;\n}\n\n/**\n * Pull the media URL out of a model's response payload.\n *\n * `path` is the dot-path from the catalog (e.g. \"images.0.url\"). Model schemas\n * drift, so if the path misses we fall back to walking the whole payload for\n * the first http(s) URL that looks like the right media type. Better to show a\n * slightly-wrong asset than a broken card.\n */\nexport function extractMediaUrl(\n payload: unknown,\n path: string,\n modality: \"image\" | \"video\" | \"audio\",\n): string | null {\n const direct = readPath(payload, path);\n if (typeof direct === \"string\" && isHttp(direct)) return direct;\n\n const wanted = MEDIA_EXTENSIONS[modality];\n const candidates: string[] = [];\n collectStrings(payload, candidates);\n\n // Prefer a URL whose extension matches the modality, then any http URL.\n const byExtension = candidates.find((u) =>\n wanted.some((ext) => u.toLowerCase().includes(ext)),\n );\n if (byExtension) return byExtension;\n\n return candidates.find(isHttp) ?? null;\n}\n\nconst MEDIA_EXTENSIONS: Record = {\n image: [\".png\", \".jpg\", \".jpeg\", \".webp\", \".gif\", \".avif\"],\n video: [\".mp4\", \".webm\", \".mov\"],\n audio: [\".mp3\", \".wav\", \".m4a\", \".ogg\", \".opus\"],\n};\n\nfunction isHttp(v: string): boolean {\n return v.startsWith(\"http://\") || v.startsWith(\"https://\");\n}\n\nfunction readPath(obj: unknown, path: string): unknown {\n let cur: unknown = obj;\n for (const key of path.split(\".\")) {\n if (cur === null || cur === undefined) return undefined;\n if (Array.isArray(cur)) {\n cur = cur[Number(key)];\n } else if (typeof cur === \"object\") {\n cur = (cur as Record)[key];\n } else {\n return undefined;\n }\n }\n return cur;\n}\n\nfunction collectStrings(node: unknown, out: string[], depth = 0): void {\n if (depth > 6 || out.length > 200) return;\n if (typeof node === \"string\") {\n if (isHttp(node)) out.push(node);\n return;\n }\n if (Array.isArray(node)) {\n for (const item of node) collectStrings(item, out, depth + 1);\n return;\n }\n if (node && typeof node === \"object\") {\n for (const value of Object.values(node)) {\n collectStrings(value, out, depth + 1);\n }\n }\n}\n\n/** Keep error messages short enough to render in the UI. */\nfunction trim(s: string, max = 400): string {\n return s.length > max ? `${s.slice(0, max)}…` : s;\n}",[148],"fa8e2948-65cf-4f25-8d08-fe4a8fd275f5","db.ts","4a7f34fa-b1e4-11f1-b367-1607ee4eb77e",8,["D",1789572984177],1,"// Project-scoped SQLite. One row per generation request, updated in place as\n// the job moves through fal's queue.\n\nimport { sqlite } from \"https://esm.town/v/std/sqlite/main.ts\";\n// Shared with the browser. It lives under frontend/ because the client fetches\n// it over HTTP; importing it here is what keeps one definition of the row shape.\nimport type { Generation, Status } from \"./frontend/types.ts\";\n\nexport type { Generation, Status };\n\nlet ready: Promise | null = null;\n\n/** Idempotent schema setup, memoised so it runs once per instance. */\nexport function initDb(): Promise {\n ready ??= (async () => {\n await sqlite.execute(`CREATE TABLE IF NOT EXISTS generations (\n id INTEGER PRIMARY KEY AUTOINCREMENT,\n modality TEXT NOT NULL,\n model TEXT NOT NULL,\n model_label TEXT NOT NULL,\n prompt TEXT NOT NULL,\n status TEXT NOT NULL DEFAULT 'queued',\n request_id TEXT,\n status_url TEXT,\n response_url TEXT,\n result_url TEXT,\n error TEXT,\n created_at TEXT NOT NULL DEFAULT (datetime('now')),\n updated_at TEXT NOT NULL DEFAULT (datetime('now'))\n )`);\n await sqlite.execute(\n `CREATE INDEX IF NOT EXISTS idx_generations_created\n ON generations (created_at DESC)`,\n );\n })();\n return ready;\n}\n\n/** Rows come back snake_case from SQLite; the UI wants camelCase. */\nfunction toGeneration(row: Record): Generation {\n return {\n id: Number(row.id),\n modality: String(row.modality),\n model: String(row.model),\n modelLabel: String(row.model_label),\n prompt: String(row.prompt),\n status: row.status as Status,\n requestId: row.request_id as string | null,\n statusUrl: row.status_url as string | null,\n responseUrl: row.response_url as string | null,\n resultUrl: row.result_url as string | null,\n error: row.error as string | null,\n createdAt: String(row.created_at),\n updatedAt: String(row.updated_at),\n };\n}\n\nexport async function insertGeneration(input: {\n modality: string;\n model: string;\n modelLabel: string;\n prompt: string;\n requestId: string;\n statusUrl: string;\n responseUrl: string;\n}): Promise {\n await initDb();\n const result = await sqlite.execute({\n sql: `INSERT INTO generations\n (modality, model, model_label, prompt, status, request_id, status_url, response_url)\n VALUES (?, ?, ?, ?, 'queued', ?, ?, ?)`,\n args: [\n input.modality,\n input.model,\n input.modelLabel,\n input.prompt,\n input.requestId,\n input.statusUrl,\n input.responseUrl,\n ],\n });\n\n const id = Number(result.lastInsertRowid);\n const row = await getGeneration(id);\n if (!row) throw new Error(\"Insert succeeded but row could not be read back\");\n return row;\n}\n\nexport async function getGeneration(id: number): Promise {\n await initDb();\n const result = await sqlite.execute({\n sql: \"SELECT * FROM generations WHERE id = ?\",\n args: [id],\n });\n const row = result.rows[0];\n return row ? toGeneration(row) : null;\n}\n\nexport async function listGenerations(limit = 60): Promise {\n await initDb();\n const result = await sqlite.execute({\n sql: \"SELECT * FROM generations ORDER BY id DESC LIMIT ?\",\n args: [limit],\n });\n return result.rows.map(toGeneration);\n}\n\nexport async function markRunning(id: number): Promise {\n await sqlite.execute({\n sql: `UPDATE generations\n SET status = 'running', updated_at = datetime('now')\n WHERE id = ? AND status = 'queued'`,\n args: [id],\n });\n}\n\nexport async function markDone(id: number, resultUrl: string): Promise {\n await sqlite.execute({\n sql: `UPDATE generations\n SET status = 'done', result_url = ?, error = NULL,\n updated_at = datetime('now')\n WHERE id = ?`,\n args: [resultUrl, id],\n });\n}\n\nexport async function markFailed(id: number, error: string): Promise {\n await sqlite.execute({\n sql: `UPDATE generations\n SET status = 'failed', error = ?, updated_at = datetime('now')\n WHERE id = ?`,\n args: [error.slice(0, 500), id],\n });\n}\n\nexport async function deleteGeneration(id: number): Promise {\n await initDb();\n await sqlite.execute({\n sql: \"DELETE FROM generations WHERE id = ?\",\n args: [id],\n });\n}",[156],"0549a5ae-6fc8-45b7-832c-817407dd4c24","README.md","01a0aada-f650-726c-8e46-8bf61b6a12f9",["D",1789572871721],"# React + Hono Starter\n\nA minimal starter template for building client-side React apps with a Hono\nbackend on Val Town. Click **Remix** to get your own copy.\n\n## Picking a starter\n\n| Starter | When to pick it |\n| --- | --- |\n| [`basic-html-starter`](https://www.val.town/x/templates/basic-html-starter) | Static HTML pages. Sprinkle of vanilla JS. No server-side rendering. |\n| [`hono-jsx-starter`](https://www.val.town/x/templates/hono-jsx-starter) | Dynamic server-rendered HTML with JSX. API routes. No React, no SPA. |\n| **[`react-hono-starter`](https://www.val.town/x/templates/react-hono-starter)** *(you're here)* | Client-side React SPA with a Hono API backend. |\n\n**Want to build UI?** Start writing React components in\n[`frontend/components/`](/frontend/components/). Import them into `App.tsx` and\nyou're good to go. Styling is [Twind](https://twind.style/) — just use Tailwind\nutility classes, zero config.\n\n**Want to add an API?** Add routes in [`index.ts`](/index.ts) (there's a\ncommented-out example). Fetch them from your React components.\n\n## Architecture\n\n```mermaid\ngraph LR\n Browser -->|\"GET /\"| Hono[\"index.ts
(Hono server)\"]\n Hono -->|serves| HTML[\"root.tsx\"]\n HTML -->|loads| Entry[\"index.tsx\"]\n Entry -->|renders| App[\"App.tsx\"]\n App -->|renders| Counter[\"Counter.tsx\"]\n```\n\n## File structure\n\n```\nindex.ts ← Hono backend (serves frontend + API routes)\nfrontend/\n root.tsx ← HTML shell (stamps immutable asset URLs)\n index.tsx ← React entrypoint (mounts )\n favicon.svg ← App icon\n components/\n App.tsx ← Root component (layout + view source link)\n Counter.tsx ← Example component with state\n```\n\n## How it works\n\n1. **`index.ts`** is the HTTP entrypoint — a Hono server that serves the\n frontend and any API routes you add.\n2. **`frontend/root.tsx`** is the HTML shell — it loads Twind and stamps\n immutable asset URLs for the favicon and React entrypoint.\n3. **`frontend/index.tsx`** mounts the React app.\n4. **`frontend/components/App.tsx`** is the root layout — add new components\n alongside it.\n5. **`frontend/components/Counter.tsx`** is an example stateful component —\n replace it with your own!\n\n## Next steps\n\n- Add API routes in `index.ts`\n- Add new components in `frontend/components/`\n- Add a SQLite database — see the\n [React Hono Example](https://www.val.town/x/stevekrouse/reactHonoExample) for\n a full-stack example with database, queries, and shared types\n\nTemplate updates and upgrade paths for existing remixes: [CHANGELOG.md](/CHANGELOG.md)\n",[164],"0416b9da-c795-4e53-a751-9cacd64df493","CHANGELOG.md","01a0aada-f650-726c-8e46-873bd723a0ba",["D",1789572871721],"# Changelog\n\n## 2026-07-30 — Immutable asset caching\n\nFrontend files are now served at version-addressed URLs\n(`/__immutable/45/frontend/index.tsx`) with\n`Cache-Control: public, max-age=31536000, immutable`, so repeat visits load the\nwhole client graph from the browser cache — roughly 3× faster (~665ms → ~157ms\nmeasured). Publishing your val bumps its version, which changes every URL, so\ncaches invalidate automatically. Old-version URLs return 404 after a publish\n(like Next.js build assets); a reload picks up the new version.\n\nWhat changed in the template:\n\n- `frontend/index.html` → [`frontend/root.tsx`](/frontend/root.tsx): the HTML\n shell is now JSX and stamps asset URLs with `immutableFileUrl()`\n- [`index.ts`](/index.ts): serves `Root()` at `/`, and one\n `app.get(\"/__immutable/*\", (c) => serveImmutableFile(c.req.path))` route\n serves all frontend files\n\n### Upgrading an existing remix\n\nYour copy of this template doesn't update automatically. Two options:\n\n**Minimal (most of the benefit):** point both routes at `serveImmutableFile` —\nyour existing bare URLs redirect into versioned space, the new route serves\nthem, and the whole module graph gets cached (costs one redirect per page view):\n\n```ts\nimport { serveImmutableFile } from \"https://esm.town/v/std/utils/index.ts\";\n\napp.get(\"/__immutable/*\", (c) => serveImmutableFile(c.req.path));\napp.get(\"/frontend/**/*\", (c) => serveImmutableFile(c.req.path)); // was serveFile\n```\n\n**Full upgrade (also removes the redirect):** three self-contained steps —\nno other files needed. Adjust the ``, favicon, and body of `Root()` to\nmatch your app's existing `index.html`, and stamp anything else your shell\nreferences directly (images, css) with `immutableFileUrl()`.\n\n1. Create `frontend/root.tsx`:\n\n```tsx\n/** @jsxImportSource npm:hono@4/jsx */\nimport { raw } from \"npm:hono@4/html\";\nimport { immutableFileUrl } from \"https://esm.town/v/std/utils/index.ts\";\n\n// The HTML shell for the app. It is never cached, so immutableFileUrl can stamp\n// the current val version into each asset URL — the browser caches those\n// immutably, and publishing the val changes the URLs, refetching exactly once.\nexport function Root() {\n return (\n <>\n {raw(\"<!DOCTYPE html>\")}\n <html lang=\"en\">\n <head>\n <meta charSet=\"UTF-8\" />\n <meta name=\"viewport\" content=\"width=device-width, initial-scale=1.0\" />\n <title>React Hono Val Town Starter\n

Diff · v16