[{"_1":2,"_3":-5,"_4":-5},"loaderData",{"_5":6,"_7":8,"_9":10},"actionData","errors","root",{"_262":263,"_264":265},"routes/_app.x.$handle.$name",{"_11":12,"_13":14,"_15":16,"_17":18,"_39":-7,"_132":-5,"_19":133,"_134":135,"_136":137,"_138":139,"_140":141,"_21":142,"_143":-5,"_144":-5,"_145":24,"_146":147,"_148":91,"_149":-5,"_150":-5,"_151":152},"routes/_app.x.$handle.$name.branch.$branch.history_.$",{"_11":12,"_13":14,"_15":16,"_17":18,"_19":20,"_21":22,"_23":24,"_25":26,"_27":28,"_29":24,"_30":-5},"handle","boh","name","ai-video-editor","branchName","main","path","main.ts","project",{"_33":79,"_13":14,"_37":80,"_81":-5,"_82":24,"_83":84,"_85":84,"_86":87,"_88":-5,"_30":66,"_89":62,"_90":91,"_92":-5,"_93":-5,"_94":26,"_41":42,"_95":96,"_97":98,"_99":100,"_101":-5,"_102":103,"_104":103,"_105":106,"_107":106,"_108":106,"_109":103,"_110":103,"_51":111,"_112":113,"_114":115,"_116":24},"file",{"_61":62,"_13":18,"_63":64,"_65":24,"_78":-5},"isDirectory",false,"page",0,"history",[31,32],"hasNext","userId",{"_33":69,"_35":36,"_37":70,"_39":71,"_41":42,"_43":72,"_44":-5,"_45":73,"_46":-5,"_47":-5,"_48":-5,"_49":-5,"_50":-5,"_51":74,"_21":75,"_54":-5,"_55":-5,"_56":-5,"_57":-5,"_58":76},{"_33":34,"_35":36,"_37":38,"_39":40,"_41":42,"_43":-5,"_44":-5,"_45":-5,"_46":-5,"_47":-5,"_48":-5,"_49":-5,"_50":-5,"_51":52,"_21":53,"_54":-5,"_55":-5,"_56":-5,"_57":-5,"_58":59},"id","99b01a64-cddb-4542-8ba3-3870917724f1","branchId","d0e0f9e5-8ba8-4507-85a6-48258752e110","createdAt",["D",1787408834177],"version",1,"orgId","01a02515-3428-7129-a58d-ee30dee1111b","parentId","incomingCommitId","aiChatMessageId","aiChatThreadId","aiChatThreadUserId","aiChatThreadTitle","source","clientLabel","user",{"_33":66,"_41":42,"_11":12,"_67":68},{"_61":62,"_13":18,"_63":64,"_65":24},"pull","merge","revert","multiple","commits",[60],{"_33":34,"_35":36,"_37":38,"_39":40,"_41":42,"_43":-5,"_44":-5,"_45":-5,"_46":-5,"_47":-5,"_48":-5,"_49":-5,"_50":-5,"_51":52,"_21":53,"_54":-5,"_55":-5,"_56":-5},"fileId","91ce646a-9e35-11f1-8e1d-1607ee4eb77e","fileType","http","deleted","86c0904a-df90-4b28-8be9-0d0cd34688de","avatar","https://lh3.googleusercontent.com/a/ACg8ocK88V05YV8HsO4yAaG3hTsa_rYvYnQ7JithXHAoBkrpvT9HmN8=s96-c","6b615af0-14d2-46da-9b9c-daeffe2e57ca",["D",1787408971613],9,"b6100482-80d0-447a-8d70-c598773e9c6e","0cfea24b-41e9-458b-b518-a0c84cabba61",{"_33":66,"_41":42,"_11":12,"_67":68},{"_61":62,"_13":18,"_63":64,"_65":24},[77],{"_33":69,"_35":36,"_37":70,"_39":71,"_41":42,"_43":72,"_44":-5,"_45":73,"_46":-5,"_47":-5,"_48":-5,"_49":-5,"_50":-5,"_51":74,"_21":75,"_54":-5,"_55":-5,"_56":-5},"deletedAtVersion","d565aca1-3a7a-4d20-8c25-06afae036bb2",["D",1787408834126],"deletedAt","legacyEnvVars","privacy","public","httpPrivacy","description","Pipeline serverless di editing video automatizzato con AI","imageUrl","pinnedFileId","pinPreview",true,"pinMigrated","screenshotVersion","likeCount","tursoName","d565aca13a7a4d208c2506afae036bb2","tursoHostname","d565aca13a7a4d208c2506afae036bb2-val-town.aws-us-east-2.turso.io","tursoMeasuredAt",["D",1789494036282],"pinned","userPerm","r","valPerm","blobPerm","rw","emailPerm","sqlitePerm","projectPerm","telemetryPerm",{"_11":12,"_131":68},"org",{"_11":12,"_126":127,"_67":68,"_30":66,"_128":129,"_130":24},"branch",{"_39":117,"_13":16,"_33":36,"_118":79,"_119":-5,"_120":-5,"_121":-5,"_122":-5,"_123":-5,"_124":14,"_125":12},"isOwned",10,"projectId","forkedBranchId","forkedBranchVersion","forkedProjectPrivacy","forkedProjectId","forkedProjectOrgId","projectName","userHandle","displayName","BCG Collaborazioni","tier","free","canRestrictVals","profileImageUrl","adoptOrgId",{"_33":79,"_13":14,"_37":257,"_81":-5,"_82":24,"_83":84,"_85":84,"_86":87,"_88":-5,"_30":66,"_89":62,"_90":91,"_92":-5,"_93":-5,"_94":26,"_41":42,"_95":96,"_97":98,"_99":258,"_101":-5,"_102":103,"_104":103,"_105":106,"_107":106,"_108":106,"_109":103,"_110":103,"_51":259,"_112":260,"_114":261,"_116":24},"tree",[171,178,179,180],"files",[171,172,173,174,175,176,177,178,179,180],"counts",{"_161":26,"_162":40,"_163":26,"_164":165,"_166":167},"contents",[],{"_33":153,"_13":18,"_61":62,"_43":-5,"_35":36,"_154":71,"_155":91,"_37":156,"_39":40,"_63":64,"_157":158,"_65":24,"_159":14,"_17":160},"upstreamProject","upstreamBranch","emptyEnvVars","flags",[],"hasHttpFiles","grants","orgs","customDomains",{},"7ea0a769-a377-491e-826c-454ed1703395","branchVersion","latest",["D",1787408971613],"content","// main.ts — API REST + webhook + media host.\n// Pipeline di editing video automatizzato con AI (Val Town).\n\nimport { Hono } from \"npm:hono\";\nimport { blob } from \"https://esm.town/v/std/blob/main.ts\";\nimport { httpEndpoint } from \"https://esm.town/v/std/utils/index.ts\";\nimport { STATUS, type Job, createJob, getJob, listJobs, updateJob } from \"./lib/db.ts\";\nimport type { Proposal } from \"./lib/director.ts\";\nimport {\n analyzeJob,\n approveJob,\n reconcile,\n jobUrl,\n outputUrlFor,\n} from \"./lib/pipeline.ts\";\n\nconst app = new Hono();\napp.onError((err) => Promise.reject(err));\n\nconst CT: Record = {\n mp4: \"video/mp4\", mov: \"video/quicktime\", webm: \"video/webm\",\n m4a: \"audio/mp4\", mp3: \"audio/mpeg\", mpeg: \"video/mpeg\", avi: \"video/x-msvideo\",\n mkv: \"video/x-matroska\",\n};\n\nfunction now(): string {\n return new Date().toISOString();\n}\n\napp.get(\"/\", (c) =>\n c.json({\n service: \"ai-video-editor\",\n version: \"1.0\",\n endpoints: [\n \"POST /jobs\",\n \"GET /jobs\",\n \"GET /jobs/:id\",\n \"GET /jobs/:id/proposal\",\n \"POST /jobs/:id/approve\",\n \"POST /jobs/:id/decline\",\n \"POST /jobs/:id/cancel\",\n \"POST /jobs/:id/retry\",\n \"GET /media/:id/source\",\n \"GET /media/:id/output\",\n \"POST /webhooks/render\",\n \"GET /health\",\n ],\n }),\n);\n\napp.get(\"/health\", (c) => c.json({ ok: true }));\n\n// Crea un job di editing. Multipart/form-data:\n// file = video (obbligatorio)\n// mode = \"assisted\" (consiglia, attende approvazione) | \"direct\" (rendi subito)\n// instructions = istruzioni di editing (assisted)\n// timeline = JSON Shotstack (direct)\napp.post(\"/jobs\", async (c) => {\n const form = await c.req.formData().catch(() => null);\n if (!form) return c.json({ error: \"multipart/form-data richiesto\" }, 400);\n\n const file = form.get(\"file\");\n if (!(file instanceof File)) return c.json({ error: \"campo 'file' mancante\" }, 400);\n\n const mode = (form.get(\"mode\") as string) || \"assisted\";\n if (mode !== \"assisted\" && mode !== \"direct\") {\n return c.json({ error: \"mode deve essere assisted|direct\" }, 400);\n }\n const instructions = (form.get(\"instructions\") as string)?.trim() || \"\";\n const notifyUrl = (form.get(\"notify_url\") as string)?.trim() || null;\n const timelineRaw = (form.get(\"timeline\") as string)?.trim();\n\n const id = crypto.randomUUID();\n const ext = (file.name.split(\".\").pop() || \"mp4\").toLowerCase();\n const sourceKey = `jobs/${id}/source.${ext}`;\n\n await blob.set(sourceKey, file);\n\n let timeline: string | null = null;\n if (mode === \"direct\") {\n if (!timelineRaw) {\n return c.json({ error: \"modalità direct richiede una 'timeline' JSON\" }, 400);\n }\n try {\n JSON.parse(timelineRaw);\n } catch {\n return c.json({ error: \"timeline non è un JSON valido\" }, 400);\n }\n timeline = timelineRaw;\n }\n\n const job: Job = {\n id, mode, instructions, sourceKey, sourceExt: ext,\n transcript: null, proposal: null, timeline,\n status: STATUS.CREATED, renderProvider: null, renderId: null,\n renderUrl: null, outputKey: null, notifyUrl, error: null,\n createdAt: now(), updatedAt: now(),\n };\n await createJob(job);\n\n if (mode === \"direct\") {\n void startRenderBackground(id);\n return c.json({ job: publicJob(job), next: null }, 201);\n }\n // assisted: analizza in background (trascrizione + proposta).\n void analyzeJob(id);\n return c.json(\n { job: publicJob(job), next: { proposal: `${jobUrl(id)}/proposal` } },\n 201,\n );\n});\n\nasync function startRenderBackground(id: string) {\n const { startRender } = await import(\"./lib/pipeline.ts\");\n await startRender(id);\n}\n\napp.get(\"/jobs\", async (c) => {\n const jobs = await listJobs(50);\n return c.json({ jobs: jobs.map(publicJob) });\n});\n\napp.get(\"/jobs/:id\", async (c) => {\n const job = await getJob(c.req.param(\"id\"));\n if (!job) return c.json({ error: \"job non trovato\" }, 404);\n return c.json({ job: publicJob(job) });\n});\n\napp.get(\"/jobs/:id/proposal\", async (c) => {\n const job = await getJob(c.req.param(\"id\"));\n if (!job) return c.json({ error: \"job non trovato\" }, 404);\n if (!job.proposal) return c.json({ error: \"proposta non disponibile\" }, 404);\n const proposal: Proposal = JSON.parse(job.proposal);\n return c.json({\n jobId: job.id,\n status: job.status,\n summary: proposal.summary,\n hooks: proposal.hooks,\n subtitles: proposal.subtitles,\n timeline: proposal.timeline,\n approveUrl: `${httpEndpoint(`/jobs/${job.id}/approve`)}`,\n declineUrl: `${httpEndpoint(`/jobs/${job.id}/decline`)}`,\n });\n});\n\napp.post(\"/jobs/:id/approve\", async (c) => {\n try {\n const job = await approveJob(c.req.param(\"id\"));\n if (!job) return c.json({ error: \"job non trovato\" }, 404);\n return c.json({ job: publicJob(job) });\n } catch (e) {\n return c.json({ error: e instanceof Error ? e.message : String(e) }, 409);\n }\n});\n\napp.post(\"/jobs/:id/decline\", async (c) => {\n const id = c.req.param(\"id\");\n const job = await getJob(id);\n if (!job) return c.json({ error: \"job non trovato\" }, 404);\n if (job.status !== STATUS.PROPOSAL) return c.json({ error: `stato ${job.status}` }, 409);\n const updated = await updateJob(id, { status: STATUS.DECLINED });\n return c.json({ job: publicJob(updated!) });\n});\n\napp.post(\"/jobs/:id/cancel\", async (c) => {\n const id = c.req.param(\"id\");\n const job = await getJob(id);\n if (!job) return c.json({ error: \"job non trovato\" }, 404);\n if ([\"queued\", \"rendering\", \"analyzing\", \"proposal\"].includes(job.status)) {\n await updateJob(id, { status: STATUS.CANCELLED });\n }\n const fresh = (await getJob(id))!;\n return c.json({ job: publicJob(fresh) });\n});\n\n// Retry: re-avvia l'analisi (assisted) o il render (direct) dopo un errore.\napp.post(\"/jobs/:id/retry\", async (c) => {\n const id = c.req.param(\"id\");\n const job = await getJob(id);\n if (!job) return c.json({ error: \"job non trovato\" }, 404);\n if (job.status !== STATUS.FAILED && job.status !== STATUS.DECLINED) {\n return c.json({ error: `stato ${job.status}: solo failed/declined sono ritentabili` }, 409);\n }\n await updateJob(id, { status: STATUS.CREATED, error: null });\n if (job.mode === \"direct\") {\n void startRenderBackground(id);\n } else {\n void analyzeJob(id);\n }\n const fresh = (await getJob(id))!;\n return c.json({ job: publicJob(fresh) });\n});\n\n// Shotstack può chiamarci alla fine del render → reconcile su richiesta.\napp.post(\"/webhooks/render\", async (c) => {\n await reconcile();\n return c.json({ ok: true });\n});\n\n// Media host: serve i byte del video sorgente salvato sul blob.\napp.get(\"/media/:id/source\", async (c) => {\n const job = await getJob(c.req.param(\"id\"));\n if (!job || !job.sourceKey) return c.json({ error: \"not found\" }, 404);\n try {\n const res = await blob.get(job.sourceKey);\n return new Response(res.body, {\n headers: {\n \"Content-Type\": CT[job.sourceExt ?? \"\"] ?? \"application/octet-stream\",\n \"Content-Disposition\": `inline; filename=\"${job.id}.${job.sourceExt}\"`,\n \"Cache-Control\": \"public, max-age=60\",\n },\n });\n } catch {\n return c.json({ error: \"not found\" }, 404);\n }\n});\n\n// Media host: serve l'output renderizzato.\napp.get(\"/media/:id/output\", async (c) => {\n const job = await getJob(c.req.param(\"id\"));\n if (!job || !job.outputKey) return c.json({ error: \"output non disponibile\" }, 404);\n try {\n const res = await blob.get(job.outputKey);\n return new Response(res.body, {\n headers: {\n \"Content-Type\": \"video/mp4\",\n \"Content-Disposition\": `inline; filename=\"${job.id}.mp4\"`,\n \"Cache-Control\": \"public, max-age=60\",\n },\n });\n } catch {\n return c.json({ error: \"output not found\" }, 404);\n }\n});\n\nfunction publicJob(j: Job) {\n return {\n id: j.id,\n mode: j.mode,\n status: j.status,\n instructions: j.instructions,\n transcript: j.transcript,\n renderProvider: j.renderProvider,\n renderUrl: j.renderUrl,\n hasProposal: !!j.proposal,\n error: j.error,\n createdAt: j.createdAt,\n updatedAt: j.updatedAt,\n links: {\n self: jobUrl(j.id),\n proposal: `${jobUrl(j.id)}/proposal`,\n approve: `${jobUrl(j.id)}/approve`,\n source: `${httpEndpoint(`/media/${j.id}/source`)}`,\n output: outputUrlFor(j.id),\n },\n };\n}\n\nexport default app.fetch;\n","label",[18],"pulls","branches","forks","envVars",{"_168":26,"_170":26},"connections",{"_168":26,"_169":26},"count","unfilledCount","emptyCount",{"_33":252,"_13":211,"_61":204,"_43":-5,"_35":36,"_154":247,"_155":91,"_37":253,"_39":26,"_63":254,"_157":-5,"_65":24,"_159":-5,"_182":255,"_17":256},{"_33":244,"_13":245,"_61":246,"_43":204,"_35":36,"_154":247,"_155":91,"_37":248,"_39":26,"_63":207,"_157":249,"_65":24,"_159":-5,"_182":250,"_17":251},{"_33":236,"_13":237,"_61":238,"_43":204,"_35":36,"_154":239,"_155":91,"_37":240,"_39":26,"_63":207,"_157":241,"_65":24,"_159":-5,"_182":242,"_17":243},{"_33":228,"_13":229,"_61":230,"_43":204,"_35":36,"_154":231,"_155":91,"_37":232,"_39":26,"_63":207,"_157":233,"_65":24,"_159":-5,"_182":234,"_17":235},{"_33":220,"_13":221,"_61":222,"_43":204,"_35":36,"_154":223,"_155":91,"_37":224,"_39":26,"_63":207,"_157":225,"_65":24,"_159":-5,"_182":226,"_17":227},{"_33":212,"_13":213,"_61":214,"_43":204,"_35":36,"_154":215,"_155":91,"_37":216,"_39":26,"_63":207,"_157":217,"_65":24,"_159":-5,"_182":218,"_17":219},{"_33":201,"_13":202,"_61":203,"_43":204,"_35":36,"_154":205,"_155":91,"_37":206,"_39":26,"_63":207,"_157":208,"_65":24,"_159":-5,"_182":209,"_17":210},{"_33":194,"_13":195,"_61":196,"_43":-5,"_35":36,"_154":117,"_155":91,"_37":197,"_39":26,"_63":21,"_157":198,"_65":24,"_159":-5,"_182":199,"_17":200},{"_33":185,"_13":186,"_61":187,"_43":-5,"_35":36,"_154":188,"_155":91,"_37":189,"_39":26,"_63":190,"_157":191,"_65":24,"_159":-5,"_182":192,"_17":193},{"_33":153,"_13":18,"_61":62,"_43":-5,"_35":36,"_154":71,"_155":91,"_37":181,"_39":40,"_63":64,"_157":158,"_65":24,"_159":14,"_182":183,"_17":184},["D",1787408971613],"children",[],[18],"f8d2be3b-c159-4c19-aa20-9c8677abaeb2","cron.ts","d9c98560-9e35-11f1-83a7-1607ee4eb77e",8,["D",1787408954943],"interval","// cron.ts — Reconcile periodico dei render attivi e retry delle analisi bloccate.\n\nimport { reconcile } from \"./lib/pipeline.ts\";\n\nexport default async function (_interval: Interval) {\n await reconcile();\n console.log(\"[cron] reconcile completato\");\n}\n",[],[186],"263129d2-48d3-41da-b293-28ea4a8a3080","README.md","f7f03728-9e35-11f1-9ef1-1607ee4eb77e",["D",1787409005526],"# AI Video Editor — Backend serverless di editing automatizzato con AI\n\nPipeline serverless su **Val Town** che orchestra l'editing video professionale tramite AI:\nupload diretto del filmato, trascrizione con Whisper, proposta di montaggio generata da un LLM\n(flusso assistito con approvazione) o rendering diretto da una timeline JSON (flusso direct),\ne consegna del video pronto via Shotstack.\n\n## Architettura\n\n```mermaid\nflowchart LR\n U[Upload multipart] --> A[POST /jobs]\n A --> B{modalità}\n B -- direct --> TL[timeline JSON]\n B -- assisted --> WH[Whisper: trascrizione]\n WH --> LLM[LLM: proposizione montaggio]\n LLM --> PR[stato proposal]\n PR -- POST /approve --> TL\n TL --> SK[Shotstack Edit API]\n SK --> REND[stato rendering]\n REND --> DONE[stato done + notifica]\n cron((cron 15m)) --> REND\n```\n\n- **`main.ts`** — router HTTP (Hono): API REST, webhook, media host.\n- **`lib/db.ts`** — state machine dei job su SQLite (project-scoped).\n- **`lib/transcribe.ts`** — Whisper via URL (Groq default, fallback OpenAI), con timestamp.\n- **`lib/director.ts`** — LLM che individua i momenti migliori e genera la timeline Shotstack.\n- **`lib/renderer.ts`** — client Shotstack (submit + poll); provider astratto per Creatomate.\n- **`lib/pipeline.ts`** — orchestrazione (analisi, render, reconcile, notifiche).\n- **`lib/notify.ts`** — notifiche Slack / Discord / Telegram / webhook generico.\n- **`cron.ts`** — reconcile dei render attivi + retry analisi impantanate.\n\n## Endpoint\n\n| Metodo | Route | Descrizione |\n|---|---|---|\n| POST | `/jobs` | Crea un job (multipart: `file`, `mode`, `instructions`, `timeline`). |\n| GET | `/jobs` | Elenca gli ultimi 50 job. |\n| GET | `/jobs/:id` | Stato del job. |\n| GET | `/jobs/:id/proposal` | Proposta di montaggio (summary, hooks, sottotitoli, timeline). |\n| POST | `/jobs/:id/approve` | Approva la proposta → avvia il rendering. |\n| POST | `/jobs/:id/decline` | Declina la proposta. |\n| POST | `/jobs/:id/cancel` | Annulla un job in coda/rendering/analisi. |\n| POST | `/jobs/:id/retry` | Re-avvia analisi/render dopo un errore. |\n| GET | `/media/:id/source` | Servirà il video sorgente (media host). |\n| GET | `/media/:id/output` | Servirà il video renderizzato. |\n| POST | `/webhooks/render` | Callback readiness → reconcile su richiesta. |\n| GET | `/health` | Health check. |\n\n## Stati del job\n\n`created → analyzing → proposal → (approve) → queued → rendering → done`\n`created → (direct) → queued → rendering → done`\n`proposal → (decline) → declined` · `… → failed` · `… → cancelled`\n\n## Setup API keys\n\nEnv var del val. Impostane almeno `GROQ_API_KEY` (trascrizione) e `SHOTSTACK_API_KEY` (render).\n\n- 👉 Aggiungi `GROQ_API_KEY` qui: https://www.val.town/x/boh/ai-video-editor/environment-variables?key=GROQ_API_KEY\n- 👉 Aggiungi `SHOTSTACK_API_KEY` qui: https://www.val.town/x/boh/ai-video-editor/environment-variables?key=SHOTSTACK_API_KEY\n- `SHOTSTACK_STAGE=` `prod` per usare credits reali (default: sandbox `/stage`, gratuita e watermarked).\n- `AI_API_KEY` / `AI_MODEL` / `AI_BASE_URL` — override per l'LLM del regista (default: key OpenAI/Groq, `gpt-4o-mini`/`llama-3.3-70b-versatile`).\n- Notifiche (optional): `SLACK_WEBHOOK_URL`, `DISCORD_WEBHOOK_URL`, `TELEGRAM_BOT_TOKEN` + `TELEGRAM_CHAT_ID`, `GENERIC_WEBHOOK_URL`.\n\n## Esempi\n\n**Upload diretto in modalità assistita** (il backup consiglia e attende approvazione):\n\n```bash\ncurl -X POST https://ai-video-editor.val.run/jobs \\\n -F \"file=@boh/export.mp4\" \\\n -F \"mode=assisted\" \\\n -F \"instructions=Mantieni solo i momenti più energici, max 60 secondi, aggiungi sottotitoli\"\n```\n\n**Verifica proposta e approva**:\n\n```bash\ncurl https://ai-video-editor.val.run/jobs//proposal\ncurl -X POST https://ai-video-editor.val.run/jobs//approve\n```\n\n**Rendering diretto da timeline JSON** (v. schema Shotstack):\n\n```bash\ncurl -X POST https://ai-video-editor.val.run/jobs \\\n -F \"file=@clip.mp4\" \\\n -F \"mode=direct\" \\\n -F 'timeline={\"timeline\":{\"background\":\"#000\",\"tracks\":[{\"clips\":[{\"asset\":{\"type\":\"video\",\"src\":\"SOURCE_URL\"},\"start\":0,\"length\":10}]}]},\"output\":{\"format\":\"mp4\",\"resolution\":\"sd\"}}'\n```\n\n> Il `src` della timeline viene sovrascritto automaticamente con l'URL del media host,\n> quindi puoi inserire un placeholder.\n\n## Limiti e note\n\n- **Blob free tier: 10 MB** — i video oltre questa soglia richiedono Pro (1 GB). Per la sandbox ok.\n- L'URL temporaneo di Shotstack **scade dopo 24 h**: la copia duplicata su blob hostato in `GET /media/:id/output` ne garantisce disponibilità.\n- Cron minimo su free tier = **15 min**: il reconcile in produzione su Pro può scendere a ogni minuto (Shotstack espone comunque webhook `serve`).\n- Non usare WebSocket: il flusso è polling + webhook.\n\n## Integrazioni bot\n\nIl val è già pronto come backend azionabile via webhook/bot (Discord/Telegram/Slack):\nla pipeline espone endpoint idempotenti e notifiche per canale. Con `boh/bubble` hai anche un\nfrontend a cui collegare la creazione job via HTTP.\n",[],[195],"f4c4ecf4-3e1c-420d-a958-e4cedf9a5a30","transcribe.ts","bd6afdea-9e35-11f1-b93f-1607ee4eb77e","c5257465-2ce4-4476-b73e-c2fd2f8430cf",3,["D",1787408907346],"script","// lib/transcribe.ts — Trascrizione Whisper via URL (Groq come default, fallback OpenAI).\n\nexport interface TranscriptSegment {\n start: number; // secondi\n end: number;\n text: string;\n}\nexport interface Transcript {\n text: string;\n segments: TranscriptSegment[];\n}\n\nconst GROQ = \"https://api.groq.com/openai/v1/audio/transcriptions\";\nconst OPENAI = \"https://api.openai.com/v1/audio/transcriptions\";\n\n/**\n * Trascrive l'audio di una sorgente remota (URL pubblico).\n * Groq `whisper-large-v3-turbo` è veloce e ha un tier gratuito.\n */\nexport async function transcribeFromUrl(\n url: string,\n): Promise {\n const groqKey = Deno.env.get(\"GROQ_API_KEY\");\n const openaiKey = Deno.env.get(\"OPENAI_API_KEY\");\n if (!groqKey && !openaiKey) {\n throw new Error(\n \"Configura GROQ_API_KEY (o OPENAI_API_KEY) per la trascrizione.\",\n );\n }\n const useGroq = !!groqKey;\n const endpoint = useGroq ? GROQ : OPENAI;\n const model = useGroq ? \"whisper-large-v3-turbo\" : \"whisper-1\";\n const key = useGroq ? groqKey! : openaiKey!;\n\n const form = new FormData();\n form.append(\"model\", model);\n form.append(\"response_format\", \"verbose_json\");\n form.append(\"timestamp_granularities[]\", \"segment\");\n form.append(\"timestamp_granularities[]\", \"word\");\n form.append(\"url\", url);\n\n const res = await fetch(endpoint, {\n method: \"POST\",\n headers: { Authorization: `Bearer ${key}` },\n body: form,\n });\n if (!res.ok) {\n throw new Error(`Trascrizione fallita (${res.status}): ${await res.text()}`);\n }\n const data = await res.json();\n\n let segments: TranscriptSegment[] = [];\n if (Array.isArray(data.segments)) {\n segments = data.segments\n .map((s: Record) => ({\n start: Number(s.start ?? 0),\n end: Number(s.end ?? 0),\n text: String(s.text ?? \"\").trim(),\n }))\n .filter((s) => s.text.length > 0);\n }\n return { text: data.text ?? \"\", segments };\n}\n",[],[211,202],"lib","ded362f3-f6c6-4182-a927-37e59774eaaf","renderer.ts","bd797d5c-9e35-11f1-b93f-1607ee4eb77e",5,["D",1787408907441],"// lib/renderer.ts — Client per l'API Shotstack (Edit API). Provider astratto:\n// la timeline JSON è cross-engine, quindi aggiungere Creatomate = nuovo client.\n\nconst STAGE = \"https://api.shotstack.io/edit/stage\"; // sandbox gratuita (watermarked)\nconst PROD = \"https://api.shotstack.io/edit/v1\"; // produzione (credits)\n\nexport interface ShotstackEdit {\n timeline: Record;\n output: Record;\n}\n\nfunction baseUrl(): string {\n return Deno.env.get(\"SHOTSTACK_STAGE\") === \"prod\" ? PROD : STAGE;\n}\n\nexport interface RenderSubmit {\n id: string;\n /** URL temporaneo (se incluso nella risposta subito). */\n url?: string;\n}\n\n/** Invia la timeline JSON a Shotstack e ritorna l'id del render. */\nexport async function submitRender(edit: ShotstackEdit): Promise {\n const key = Deno.env.get(\"SHOTSTACK_API_KEY\");\n if (!key) {\n throw new Error(\"Configura SHOTSTACK_API_KEY per il rendering.\");\n }\n const res = await fetch(`${baseUrl()}/render`, {\n method: \"POST\",\n headers: {\n \"Content-Type\": \"application/json\",\n \"x-api-key\": key,\n \"Accept\": \"application/json\",\n },\n body: JSON.stringify({ timeline: edit.timeline, output: edit.output }),\n });\n const data = await res.json();\n if (!res.ok || !data?.success) {\n throw new Error(`Shotstack render fallito (${res.status}): ${JSON.stringify(data)}`);\n }\n return { id: String(data.response.id) };\n}\n\nexport interface RenderStatus {\n status: \"queued\" | \"fetching\" | \"rendering\" | \"done\" | \"failed\" | string;\n /** URL temporaneo del video finito (disponibile quando done). */\n url?: string;\n}\n\nexport async function renderStatus(id: string): Promise {\n const key = Deno.env.get(\"SHOTSTACK_API_KEY\");\n if (!key) throw new Error(\"Configura SHOTSTACK_API_KEY.\");\n const res = await fetch(`${baseUrl()}/render/${id}`, {\n headers: { \"x-api-key\": key, \"Accept\": \"application/json\" },\n });\n const data = await res.json();\n if (!res.ok || !data?.success) {\n throw new Error(`Status render fallito (${res.status}): ${JSON.stringify(data)}`);\n }\n return {\n status: String(data.response.status),\n url: data.response.url ?? undefined,\n };\n}\n",[],[211,213],"25f5627e-4364-4fd5-ac91-b726bb22b910","pipeline.ts","d9c2632a-9e35-11f1-83a7-1607ee4eb77e",7,["D",1787408954894],"// lib/pipeline.ts — Orchestrazione: analisi trascrizione+proposta, avvio render,\n// reconcile degli stati. Usato sia dal server HTTP sia dal cron.\n\nimport { STATUS, getJob, updateJob, listJobsByStatus, type Job } from \"./db.ts\";\nimport { transcribeFromUrl } from \"./transcribe.ts\";\nimport { buildProposal } from \"./director.ts\";\nimport { submitRender, renderStatus } from \"./renderer.ts\";\nimport { notify } from \"./notify.ts\";\nimport { blob } from \"https://esm.town/v/std/blob/main.ts\";\nimport { httpEndpoint } from \"https://esm.town/v/std/utils/index.ts\";\n\n/** URL pubblico dell'endpoint /media/:id/source (il val fa da media host). */\nexport function sourceUrlFor(id: string): string {\n return httpEndpoint(`/media/${id}/source`);\n}\nexport function outputUrlFor(id: string): string {\n return httpEndpoint(`/media/${id}/output`);\n}\nexport function jobUrl(id: string): string {\n return httpEndpoint(`/jobs/${id}`);\n}\n\n/** Modalità assistita: trascrizione + proposta → stato PROPOSAL. */\nexport async function analyzeJob(id: string): Promise {\n const job = await getJob(id);\n if (!job || job.status !== STATUS.CREATED) return;\n\n await updateJob(id, { status: STATUS.ANALYZING, error: null });\n try {\n const src = sourceUrlFor(id);\n const transcript = await transcribeFromUrl(src);\n const proposal = await buildProposal({\n sourceUrl: src,\n transcript,\n instructions: job.instructions ?? \"\",\n });\n await updateJob(id, {\n status: STATUS.PROPOSAL,\n transcript: transcript.text,\n proposal: JSON.stringify(proposal),\n });\n await notify({\n title: `🧠 Proposta pronta — job ${id}`,\n body: proposal.summary || \"Una proposta di montaggio è pronta per la revisione.\",\n links: {\n \"🎬 Proposta\": `${jobUrl(id)}/proposal`,\n \"✅ Approva\": `${httpEndpoint(`/jobs/${id}/approve`)}`,\n \"❌ Declina\": `${httpEndpoint(`/jobs/${id}/decline`)}`,\n },\n });\n } catch (e) {\n const msg = e instanceof Error ? e.message : String(e);\n console.error(\"[analyze]\", id, msg);\n await updateJob(id, { status: STATUS.FAILED, error: msg });\n await notify({\n title: `⚠️ Analisi fallita — job ${id}`,\n body: msg,\n links: { \"Job\": jobUrl(id) },\n });\n }\n}\n\n/** Avvia il rendering di una timeline esplicita o della proposta approvata. */\nexport async function startRender(id: string): Promise {\n const job = await getJob(id);\n if (!job) return null;\n\n const timelineJson =\n job.timeline ?? (job.proposal ? JSON.parse(job.proposal).timeline : null);\n if (!timelineJson) {\n await updateJob(id, { status: STATUS.FAILED, error: \"Nessuna timeline disponibile.\" });\n return job;\n }\n\n try {\n await updateJob(id, { status: STATUS.QUEUED, renderProvider: \"shotstack\", renderId: null, renderUrl: null, error: null });\n const { id: renderId } = await submitRender(timelineJson);\n await updateJob(id, { status: STATUS.RENDERING, renderId, renderProvider: \"shotstack\" });\n await notify({\n title: `🎬 Rendering avviato — job ${id}`,\n body: \"La timeline è in coda su Shotstack.\",\n links: { \"Job\": jobUrl(id) },\n });\n return job;\n } catch (e) {\n const msg = e instanceof Error ? e.message : String(e);\n console.error(\"[render]\", id, msg);\n await updateJob(id, { status: STATUS.FAILED, error: msg });\n await notify({\n title: `⚠️ Avvio render fallito — job ${id}`,\n body: msg,\n links: { \"Job\": jobUrl(id) },\n });\n return job;\n }\n}\n\n/** Approva la proposta: la timeline proposta passa al render. */\nexport async function approveJob(id: string): Promise {\n const job = await getJob(id);\n if (!job) return null;\n if (job.status !== STATUS.PROPOSAL) {\n throw new Error(`Job ${id} non è in stato di proposta (${job.status}).`);\n }\n return startRender(id);\n}\n\n/** Poll dei render attivi + retry delle analisi impantanate. */\nexport async function reconcile(): Promise {\n const active = [\n ...(await listJobsByStatus(STATUS.QUEUED)),\n ...(await listJobsByStatus(STATUS.RENDERING)),\n ];\n\n for (const job of active) {\n if (!job.renderId) continue;\n try {\n const st = await renderStatus(job.renderId);\n if (st.status === \"done\" && st.url) {\n await completeRender(job, st.url);\n } else if (st.status === \"failed\") {\n await updateJob(job.id, { status: STATUS.FAILED, error: \"Render Shotstack fallito.\" });\n await notify({ title: `💥 Render fallito — ${job.id}`, body: \"Lo stato su Shotstack è 'failed'.\", links: { \"Job\": jobUrl(job.id) } });\n }\n // queued/fetching/rendering → aspetta il prossimo giro.\n } catch (e) {\n console.error(\"[reconcile]\", job.id, e);\n }\n }\n\n // Retry analisi che si sono impantanate (> 10 min in ANALYZING).\n const staleAnalyzing = (await listJobsByStatus(STATUS.ANALYZING)).filter(\n (j) => Date.now() - new Date(j.updatedAt).getTime() > 10 * 60 * 1000,\n );\n for (const j of staleAnalyzing) {\n await updateJob(j.id, { status: STATUS.CREATED });\n void analyzeJob(j.id);\n }\n}\n\n/** Render completato: salva l'output su blob e passa a DONE. */\nasync function completeRender(job: Job, tempUrl: string): Promise {\n try {\n const res = await fetch(tempUrl);\n if (!res.ok) throw new Error(`Download output ${res.status}`);\n const outBlob = await res.blob();\n const key = `jobs/${job.id}/output.mp4`;\n await blob.set(key, outBlob);\n await updateJob(job.id, {\n status: STATUS.DONE,\n renderUrl: tempUrl,\n outputKey: key,\n });\n await notify({\n title: `✅ Video pronto — job ${job.id}`,\n body: \"Il render è completato.\",\n links: {\n \"▶️ Download\": tempUrl,\n \"🍿 Output (val host)\": outputUrlFor(job.id),\n },\n });\n } catch (e) {\n console.error(\"[completeRender]\", job.id, e);\n await updateJob(job.id, {\n status: STATUS.RENDERED,\n renderUrl: tempUrl,\n error: e instanceof Error ? e.message : String(e),\n });\n await notify({\n title: `✅ Render pronto (ma archiviazione fallita) — ${job.id}`,\n body: \"Scarica dal link temporaneo (scade in 24h).\",\n links: { \"▶️ Download\": tempUrl },\n });\n }\n}\n",[],[211,221],"13ee1032-e75e-4885-a9c8-39abca551c49","notify.ts","bd80f7da-9e35-11f1-b93f-1607ee4eb77e",6,["D",1787408907489],"// lib/notify.ts — Notifiche a più canali via webhook (Slack / Discord / Telegram\n// / webhook HTTP generico). Ogni canale è attivato dalla presenza della sua env var.\n\ninterface NotifyPayload {\n title: string;\n body: string;\n links: Record;\n}\n\nfunction buildText(p: NotifyPayload, prefix = \"\"): string {\n return [\n `${prefix}*${p.title}*`,\n p.body,\n ...Object.entries(p.links).map(([k, v]) => `${k}: ${v}`),\n ].join(\"\\n\");\n}\n\nasync function post(url: string, headers: Record, body: unknown) {\n try {\n await fetch(url, {\n method: \"POST\",\n headers: { \"Content-Type\": \"application/json\", ...headers },\n body: JSON.stringify(body),\n });\n } catch (e) {\n console.error(\"[notify]\", e);\n }\n}\n\nexport async function notify(p: NotifyPayload): Promise {\n const jobs: Promise[] = [];\n\n const slack = Deno.env.get(\"SLACK_WEBHOOK_URL\");\n if (slack) {\n jobs.push(\n post(slack, {}, { text: buildText(p) }),\n );\n }\n\n const discord = Deno.env.get(\"DISCORD_WEBHOOK_URL\");\n if (discord) {\n jobs.push(\n post(discord, {}, { content: buildText(p) }),\n );\n }\n\n const tBot = Deno.env.get(\"TELEGRAM_BOT_TOKEN\");\n const tChat = Deno.env.get(\"TELEGRAM_CHAT_ID\");\n if (tBot && tChat) {\n jobs.push(\n post(\n `https://api.telegram.org/bot${tBot}/sendMessage`,\n {},\n { chat_id: tChat, text: buildText(p), parse_mode: \"Markdown\" },\n ),\n );\n }\n\n const generic = Deno.env.get(\"GENERIC_WEBHOOK_URL\");\n if (generic) {\n jobs.push(post(generic, {}, { title: p.title, body: p.body, links: p.links }));\n }\n\n await Promise.allSettled(jobs);\n}\n",[],[211,229],"55701016-d7c9-45c2-b5bc-fcb4f431557c","director.ts","bd727c00-9e35-11f1-b93f-1607ee4eb77e",4,["D",1787408907395],"// lib/director.ts — \"Regista AI\": dal transcript + istruzioni produce una proposta\n// di montaggio (momenti migliori + timeline Shotstack JSON) pronta per approvazione.\n\nimport type { Transcript } from \"./transcribe.ts\";\n\nexport interface Proposal {\n summary: string;\n /** Momenti migliori individuati automaticamente. */\n hooks: { start: number; end: number; why: string }[];\n /** Sottotitoli temporizzati, generati dal transcript. */\n subtitles: { start: number; end: number; text: string }[];\n /** Timeline Shotstack pronta per il rendering. */\n timeline: Record;\n}\n\nexport interface DirectorInput {\n sourceUrl: string;\n transcript: Transcript;\n instructions: string;\n /** True per \"si immerge\" dentro le istruzioni; altrimenti decisione libera. */\n}\n\nfunction buildShotstackEdit(\n sourceUrl: string,\n subtitles: { start: number; end: number; text: string }[],\n hooks: { start: number; end: number; why: string }[],\n): Record {\n const total = Math.max(\n 2,\n ...subtitles.map((s) => s.end),\n ...hooks.map((h) => h.end),\n );\n\n // Traccia di base: l'intero video.\n const baseClips = [\n {\n asset: { type: \"video\", src: sourceUrl },\n start: 0,\n length: total,\n },\n ];\n\n // Traccia sottotitoli: un clip \"title\" per ogni segmento parlato (con un tetto).\n // Il format \"subtitle\" di Shotstack non esiste: usiamo \"title\" base e stile testo.\n const captionClips = subtitles.slice(0, 40).map((s) => ({\n asset: {\n type: \"title\",\n text: s.text,\n style: \"compress\",\n size: \"small\",\n },\n start: s.start,\n length: Math.max(0.4, s.end - s.start),\n }));\n\n return {\n timeline: {\n background: \"#000000\",\n tracks: [ { clips: baseClips }, { clips: captionClips, offset: 0 }],\n },\n output: { format: \"mp4\", resolution: \"sd\" },\n };\n}\n\nasync function callLLM(system: string, user: string): Promise {\n const apiKey =\n Deno.env.get(\"AI_API_KEY\") ??\n Deno.env.get(\"OPENAI_API_KEY\") ??\n Deno.env.get(\"GROQ_API_KEY\");\n if (!apiKey) {\n throw new Error(\n \"Configura AI_API_KEY (o OPENAI_API_KEY / GROQ_API_KEY) per l'orchestrazione AI.\",\n );\n }\n const base = Deno.env.get(\"AI_BASE_URL\") ??\n (Deno.env.get(\"GROQ_API_KEY\") ? \"https://api.groq.com/openai/v1\" : \"https://api.openai.com/v1\");\n const model = Deno.env.get(\"AI_MODEL\") ??\n (base.includes(\"groq\") ? \"llama-3.3-70b-versatile\" : \"gpt-4o-mini\");\n\n const res = await fetch(`${base}/chat/completions`, {\n method: \"POST\",\n headers: { \"Content-Type\": \"application/json\", Authorization: `Bearer ${apiKey}` },\n body: JSON.stringify({\n model,\n response_format: { type: \"json_object\" },\n temperature: 0.4,\n messages: [\n { role: \"system\", content: system },\n { role: \"user\", content: user },\n ],\n }),\n });\n if (!res.ok) {\n throw new Error(`LLM fallito (${res.status}): ${await res.text()}`);\n }\n const data = await res.json();\n const content = data?.choices?.[0]?.message?.content;\n if (!content) throw new Error(\"LLM: risposta vuota.\");\n return JSON.parse(content);\n}\n\nconst SYSTEM_PROMPT = `Sei un direttore di montaggio video esperto. Analizzi una trascrizione\ncon timestamp e produci UNA PROPOSTA DI MONTAGGIO. Devi:\n- individuare i momenti più significativi (hook), indicando perché funzionano;\n- proporre una timeline valida per l'API Shotstack (JSON: timeline + output);\n- generare sottotitoli temporizzati.\nRispondi SOLO con un JSON con questo schema:\n{\n \"summary\": \"stringa riassuntiva della proposta\",\n \"hooks\": [ { \"start\": number, \"end\": number, \"why\": \"string\" } ],\n \"subtitles\": [ { \"start\": number, \"end\": number, \"text\": \"string\" } ]\n}\nstart/end sono in secondi. Se non ci sono contenuti idonei, imposta hooks=[] e una\ntimeline minimale. Non inventare fatti non presenti nella trascrizione.`;\n\nexport async function buildProposal(\n input: DirectorInput,\n): Promise {\n const user = `ISTRUZIONI DELL'UTENTE:\\n\"\"\"${input.instructions || \"(nessuna istruzione specifica, decidi tu)\"}\"\"\"\\n\\n` +\n `TRASCRIZIONE (con timestamp):\\n` +\n input.transcript.segments\n .map((s) => `[${s.start.toFixed(1)}-${s.end.toFixed(1)}s] ${s.text}`)\n .join(\"\\n\");\n\n // 1) L'LLM sceglie i momenti migliori + sottotitoli.\n const decided = await callLLM(SYSTEM_PROMPT, user);\n const hooks: { start: number; end: number; why: string }[] = (\n Array.isArray(decided.hooks) ? decided.hooks : []\n ).map((h) => ({\n start: Number(h.start ?? 0),\n end: Number(h.end ?? 0),\n why: String(h.why ?? \"\"),\n }));\n const subtitles = (\n Array.isArray(decided.subtitles) && decided.subtitles.length\n ? decided.subtitles\n : input.transcript.segments\n ).map((s) => ({\n start: Number(s.start ?? 0),\n end: Number(s.end ?? 0),\n text: String(s.text ?? \"\"),\n }));\n\n // 2) Costruiamo materialmente la timeline Shotstack.\n const timeline = buildShotstackEdit(input.sourceUrl, subtitles, hooks);\n\n return {\n summary: String(decided.summary ?? \"\"),\n hooks,\n subtitles,\n timeline,\n };\n}\n\nexport { buildShotstackEdit };\n",[],[211,237],"b08c580c-776b-4428-8d1c-ae8a2105e2f1","db.ts","bd62ea24-9e35-11f1-b93f-1607ee4eb77e",2,["D",1787408907290],"// lib/db.ts — Modello dati e state machine dei job su SQLite (project-scoped).\n\nimport { sqlite } from \"https://esm.town/v/std/sqlite/main.ts\";\n\n/** Stati della pipeline. Flusso:\n * created ──assistita──► analyzing ─► proposal (in attesa di approvazione)\n * created ──diretta────► queued ─► rendering ─► rendered ─► done\n * proposal ─approve/decline─► queued / declined\n */\nexport const STATUS = {\n CREATED: \"created\",\n ANALYZING: \"analyzing\",\n PROPOSAL: \"proposal\", // in attesa di approvazione umana\n DECLINED: \"declined\",\n QUEUED: \"queued\",\n RENDERING: \"rendering\",\n RENDERED: \"rendered\",\n DONE: \"done\",\n FAILED: \"failed\",\n CANCELLED: \"cancelled\",\n} as const;\nexport type JobStatus = (typeof STATUS)[keyof typeof STATUS];\n\nexport type JobMode = \"assisted\" | \"direct\";\n\nexport interface Job {\n id: string;\n mode: JobMode;\n instructions: string | null;\n sourceKey: string | null;\n sourceExt: string | null;\n transcript: string | null;\n proposal: string | null; // JSON: { summary, hooks, subtitles, timeline }\n timeline: string | null; // JSON: Shotstack edit\n status: JobStatus;\n renderProvider: string | null;\n renderId: string | null;\n renderUrl: string | null;\n outputKey: string | null;\n notifyUrl: string | null;\n error: string | null;\n createdAt: string;\n updatedAt: string;\n}\n\nlet inited = false;\nasync function ensureInit() {\n if (inited) return;\n await sqlite.execute(`CREATE TABLE IF NOT EXISTS jobs (\n id TEXT PRIMARY KEY,\n mode TEXT NOT NULL,\n instructions TEXT,\n source_key TEXT,\n source_ext TEXT,\n transcript TEXT,\n proposal TEXT,\n timeline TEXT,\n status TEXT NOT NULL,\n render_provider TEXT,\n render_id TEXT,\n render_url TEXT,\n output_key TEXT,\n notify_url TEXT,\n error TEXT,\n created_at TEXT NOT NULL,\n updated_at TEXT NOT NULL\n )`);\n inited = true;\n}\n\nfunction rowToJob(row: Record | undefined): Job | null {\n if (!row) return null;\n return {\n id: row.id as string,\n mode: row.mode as JobMode,\n instructions: (row.instructions as string) ?? null,\n sourceKey: (row.source_key as string) ?? null,\n sourceExt: (row.source_ext as string) ?? null,\n transcript: (row.transcript as string) ?? null,\n proposal: (row.proposal as string) ?? null,\n timeline: (row.timeline as string) ?? null,\n status: row.status as JobStatus,\n renderProvider: (row.render_provider as string) ?? null,\n renderId: (row.render_id as string) ?? null,\n renderUrl: (row.render_url as string) ?? null,\n outputKey: (row.output_key as string) ?? null,\n notifyUrl: (row.notify_url as string) ?? null,\n error: (row.error as string) ?? null,\n createdAt: row.created_at as string,\n updatedAt: row.updated_at as string,\n };\n}\n\nexport async function createJob(j: Job): Promise {\n await ensureInit();\n await sqlite.execute({\n sql: `INSERT INTO jobs (id, mode, instructions, source_key, source_ext,\n transcript, proposal, timeline, status, render_provider, render_id,\n render_url, output_key, notify_url, error, created_at, updated_at)\n VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?)`,\n args: [\n j.id, j.mode, j.instructions, j.sourceKey, j.sourceExt,\n j.transcript, j.proposal, j.timeline, j.status, j.renderProvider, j.renderId,\n j.renderUrl, j.outputKey, j.notifyUrl, j.error, j.createdAt, j.updatedAt,\n ],\n });\n}\n\ntype Update = Partial>;\nconst COL: Record = {\n mode: \"mode\", instructions: \"instructions\", sourceKey: \"source_key\",\n sourceExt: \"source_ext\", transcript: \"transcript\", proposal: \"proposal\",\n timeline: \"timeline\", status: \"status\", renderProvider: \"render_provider\",\n renderId: \"render_id\", renderUrl: \"render_url\", outputKey: \"output_key\",\n notifyUrl: \"notify_url\", error: \"error\", updatedAt: \"updated_at\",\n};\n\nexport async function updateJob(id: string, patch: Update): Promise {\n await ensureInit();\n const sets: string[] = [];\n const args: unknown[] = [];\n for (const [k, v] of Object.entries(patch)) {\n const col = COL[k];\n if (!col) continue;\n sets.push(`${col} = ?`);\n args.push(v ?? null);\n }\n if (sets.length === 0) return getJob(id);\n sets.push(\"updated_at = ?\");\n args.push(new Date().toISOString());\n args.push(id);\n await sqlite.execute({ sql: `UPDATE jobs SET ${sets.join(\", \")} WHERE id = ?`, args });\n return getJob(id);\n}\n\nexport async function getJob(id: string): Promise {\n await ensureInit();\n const r = await sqlite.execute({ sql: \"SELECT * FROM jobs WHERE id = ?\", args: [id] });\n return rowToJob(r.rows[0] as Record | undefined);\n}\n\nexport async function listJobsByStatus(status: JobStatus): Promise {\n await ensureInit();\n const r = await sqlite.execute({ sql: \"SELECT * FROM jobs WHERE status = ? ORDER BY created_at ASC\", args: [status] });\n return (r.rows as Record[]).map(rowToJob).filter((x): x is Job => !!x);\n}\n\nexport async function listJobs(limit = 50): Promise {\n await ensureInit();\n const r = await sqlite.execute({ sql: \"SELECT * FROM jobs ORDER BY created_at DESC LIMIT ?\", args: [limit] });\n return (r.rows as Record[]).map(rowToJob).filter((x): x is Job => !!x);\n}\n",[],[211,245],"bc884ed6-cf89-4e01-b5a7-5c518d0390d5",["D",1787408907290],"directory",[172,173,174,175,176,177],[211],["D",1787408834126],["D",1789494036282],{"_11":12,"_131":68},{"_11":12,"_126":127,"_67":68,"_30":66,"_128":129,"_130":24},{"_39":117,"_13":16,"_33":36,"_118":79,"_119":-5,"_120":-5,"_121":-5,"_122":-5,"_123":-5,"_124":14,"_125":12},"ctx",{"_51":-5,"_150":295,"_296":297,"_298":299,"_300":24,"_301":24,"_302":24,"_303":91,"_304":-5,"_305":-7,"_306":-7,"_307":26,"_308":24,"_309":24,"_310":-5,"_311":24,"_312":24,"_313":24,"_314":-5},"ENV",{"_266":267,"_268":269,"_270":16,"_271":272,"_273":274,"_275":276,"_277":278,"_279":280,"_281":282,"_283":284,"_285":286,"_287":288,"_289":290,"_291":292,"_293":294},"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"]

History · main.ts