[{"data":1,"prerenderedAt":2218},["ShallowReactive",2],{"infra-ecosystem":3},[4,26,42,58,229,830,1024,1184,1580,1763,1983],{"id":5,"title":6,"body":7,"contentId":14,"description":15,"ecosystem":16,"extension":17,"meta":18,"navigation":19,"part":16,"path":20,"rawbody":21,"repository":16,"robots":16,"schemaOrg":16,"seo":22,"sitemap":23,"stem":24,"url":16,"__hash__":25},"pages\u002Finfra\u002Finternal-tools\u002Findex.md","Internal tools",{"type":8,"value":9,"toc":10},"minimark",[],{"title":11,"searchDepth":12,"depth":12,"links":13},"",2,[],"d6c94db9-c150-4777-b9f6-1577d3bfdcbb","The private tools I use to manage my infrastructure.",null,"md",{},true,"\u002Finfra\u002Finternal-tools","---\ncontentId: d6c94db9-c150-4777-b9f6-1577d3bfdcbb\ntitle: Internal tools\ndescription: The private tools I use to manage my infrastructure.\n---\n",{"title":6,"description":15},{"loc":20},"infra\u002Finternal-tools\u002Findex","_4Eq1nyToweC3mV1aRGcLu8RymN1yij64Nr0jQbxfs8",{"id":27,"title":28,"body":29,"contentId":33,"description":34,"ecosystem":16,"extension":17,"meta":35,"navigation":19,"part":16,"path":36,"rawbody":37,"repository":16,"robots":16,"schemaOrg":16,"seo":38,"sitemap":39,"stem":40,"url":16,"__hash__":41},"pages\u002Finfra\u002Fservices\u002Findex.md","Services",{"type":8,"value":30,"toc":31},[],{"title":11,"searchDepth":12,"depth":12,"links":32},[],"7e967564-582f-499c-8ec3-fca15efd1af4","The services that support my websites and infrastructure.",{},"\u002Finfra\u002Fservices","---\ncontentId: 7e967564-582f-499c-8ec3-fca15efd1af4\ntitle: Services\ndescription: The services that support my websites and infrastructure.\n---\n",{"title":28,"description":34},{"loc":36},"infra\u002Fservices\u002Findex","-mE8RPD0qZpXUsIRQsj_q3UBecUFOqFozuKIQIr5mgc",{"id":43,"title":44,"body":45,"contentId":49,"description":50,"ecosystem":16,"extension":17,"meta":51,"navigation":19,"part":16,"path":52,"rawbody":53,"repository":16,"robots":16,"schemaOrg":16,"seo":54,"sitemap":55,"stem":56,"url":16,"__hash__":57},"pages\u002Finfra\u002Fwebsites\u002Findex.md","Websites",{"type":8,"value":46,"toc":47},[],{"title":11,"searchDepth":12,"depth":12,"links":48},[],"0497b82d-fdf9-4427-beca-c1002186cbfc","The public websites I build and maintain.",{},"\u002Finfra\u002Fwebsites","---\ncontentId: 0497b82d-fdf9-4427-beca-c1002186cbfc\ntitle: Websites\ndescription: The public websites I build and maintain.\n---\n",{"title":44,"description":50},{"loc":52},"infra\u002Fwebsites\u002Findex","Z9CRAN2G7QVNdnSR9xvvGXEDKzO478XOH8Mrk2RNScU",{"id":59,"title":60,"body":61,"contentId":169,"description":170,"ecosystem":171,"extension":17,"meta":222,"navigation":19,"part":16,"path":223,"rawbody":224,"repository":181,"robots":16,"schemaOrg":16,"seo":225,"sitemap":226,"stem":227,"url":69,"__hash__":228},"pages\u002Finfra\u002Fservices\u002Fapi-soubiran-dev.md","api.soubiran.dev",{"type":8,"value":62,"toc":163},[63,78,89,94,97,100,103,107,123,126,129,132,135,138,142,145,153,156,160],[64,65,66,72,73,77],"p",{},[67,68,60],"a",{"href":69,"rel":70},"https:\u002F\u002Fapi.soubiran.dev",[71],"nofollow"," owns the dynamic data behind ",[67,74,76],{"href":75},"\u002Finfra\u002Fwebsites\u002Fsoubiran-dev","soubiran.dev",". The personal website remains statically generated, while this Laravel application stores user accounts, comments, reactions, votes, feedback, messages, and notifications.",[64,79,80,81,83,84,88],{},"It also imports stable records for content published by ",[67,82,76],{"href":75},", including its ",[67,85,87],{"href":86},"\u002Finfra","infrastructure documentation",". Stable page IDs keep those records connected when a title or URL changes.",[90,91,93],"h2",{"id":92},"why-the-dynamic-state-is-separate","Why the dynamic state is separate",[64,95,96],{},"Most requests to my personal website only need static HTML and assets. Comments and reactions have different requirements. They need authenticated writes, authorization, persistence, and an administration interface. Putting those concerns in a separate application keeps them out of the static publishing path.",[64,98,99],{},"Laravel provides the data model and request lifecycle for that mutable state. The website can still deploy as static files on Cloudflare, and its browser code calls the API only for interactive features. This split fits a content site that benefits from static delivery but still needs accounts and user-generated data.",[64,101,102],{},"The API stores a local page record rather than copying article content. Published records use the UUID from the source website as their API route key. Interactions point to that record, so a later content build can change the page title or canonical URL without breaking its comments or reactions.",[90,104,106],{"id":105},"how-content-and-interactions-meet","How content and interactions meet",[64,108,109,110,114,115,118,119,122],{},"The website publishes ",[111,112,113],"code",{},"\u002Fpages.en.json"," and ",[111,116,117],{},"\u002Fpages.fr.json"," catalogs. During deployment, the API fetches those catalogs and upserts their entries by UUID. The ",[67,120,121],{"href":75},"soubiran.dev catalogs"," supply each page's stable ID, title, description, published URL, and Markdown URL. The English catalog also includes the infrastructure documentation.",[64,124,125],{},"The browser addresses published-page interaction routes with that UUID. Public endpoints return emoji definitions, vote totals, feedback summaries, reactions, and comment threads. They also accept page feedback without creating an account. Comment responses include rendered Markdown, replies, like counts, author data, and capability flags, which lets the website render the thread without reproducing Laravel's authorization rules.",[64,127,128],{},"Voting is the exception to UUID routing. A candidate article can exist in the API before publication, so vote mutations use its internal database ID. Once the article is published, a metadata import attaches the public UUID to the same record.",[64,130,131],{},"Authenticated routes cover user-specific data and mutations. They return the current account, votes, reactions, and notifications, and they handle comments, likes, reactions, votes, Markdown previews, and ask-me-anything messages. The API only includes private identity fields when a signed-in user reads their own account.",[64,133,134],{},"Laravel Sanctum treats the website as a stateful first-party client. Before an unsafe request, the browser obtains a CSRF cookie and sends the request with its session credentials. Users sign in through GitHub, Google, or Discord with Laravel Socialite. This uses the browser session for the first-party website instead of exposing long-lived API tokens to client code.",[64,136,137],{},"Filament provides the private administration panel. New accounts and user activity can create queued database notifications for administrators, while Laravel policies decide which comment and account actions each request may perform.",[90,139,141],{"id":140},"deployment-ordering","Deployment ordering",[64,143,144],{},"Forge deploys the application on a server. Its deployment script enables maintenance mode, pulls the selected branch, installs production PHP dependencies, optimizes Laravel, reloads PHP-FPM, and runs database migrations. It then imports the website's localized catalogs before bringing the application back online.",[64,146,147,148,152],{},"Those imports make deployment order part of the data contract. A normal API deployment must fetch catalogs that are already live, not files from the previous website build. After a push to the website, ",[67,149,151],{"href":150},"\u002Finfra\u002Fservices\u002Fredeploy-soubiran-dev","redeploy.soubiran.dev"," waits for the corresponding Cloudflare Worker deployment to reach production before it calls the API deploy hook.",[64,154,155],{},"This sequence keeps static content and mutable records independent without letting their shared page IDs drift. Website builds publish the source metadata first. The API deployment imports it second. Browser requests can then attach current interaction data to the new pages.",[90,157,159],{"id":158},"ecosystem","Ecosystem",[161,162],"infra-ecosystem",{},{"title":11,"searchDepth":12,"depth":12,"links":164},[165,166,167,168],{"id":92,"depth":12,"text":93},{"id":105,"depth":12,"text":106},{"id":140,"depth":12,"text":141},{"id":158,"depth":12,"text":159},"f2225f4f-7da6-4515-a5bf-e6adf1c9f294","The Laravel service that owns accounts, comments, reactions, votes, feedback, and other dynamic data for my personal website.",[172,211,219],{"type":173,"name":174,"description":175,"ecosystem":176},"deployment","Forge","Deploy the Laravel application and run its database migrations and metadata imports.",[177,208],{"type":178,"id":60,"name":179,"description":180,"href":181,"ecosystem":182},"repository","GitHub","Host the public application source and Forge deployment script.","https:\u002F\u002Fgithub.com\u002Fbarbapapazes\u002Fapi.soubiran.dev",[183,188,192,196,200,204],{"type":184,"name":185,"description":186,"href":187},"stack","Laravel","Handle HTTP routes, persistence, authorization, queues, and notifications.","https:\u002F\u002Flaravel.com",{"type":184,"name":189,"description":190,"href":191},"Laravel Sanctum","Authenticate first-party browser requests with the Laravel session.","https:\u002F\u002Flaravel.com\u002Fdocs\u002Fsanctum",{"type":184,"name":193,"description":194,"href":195},"Laravel Socialite","Connect user accounts to GitHub, Google, and Discord.","https:\u002F\u002Flaravel.com\u002Fdocs\u002Fsocialite",{"type":184,"name":197,"description":198,"href":199},"Filament","Provide the private administration panel for application records.","https:\u002F\u002Ffilamentphp.com",{"type":184,"name":201,"description":202,"href":203},"Spatie Laravel Markdown","Render the Markdown preview used when writing comments.","https:\u002F\u002Fgithub.com\u002Fspatie\u002Flaravel-markdown",{"type":184,"name":205,"description":206,"href":207},"Sentry","Capture application exceptions and performance traces.","https:\u002F\u002Fsentry.io",{"type":209,"name":151,"description":210},"service","Trigger this deployment only after an upstream website reaches production.",{"type":212,"name":213,"description":214,"ecosystem":215},"data","soubiran.dev metadata","Supply stable page IDs, titles, and URLs for personal website content and infrastructure documentation.",[216],{"type":217,"name":76,"description":218},"website","Publish the metadata during its static build.",{"type":217,"relationship":220,"name":76,"description":221},"consumer","Read public interaction data and send authenticated changes from the browser.",{},"\u002Finfra\u002Fservices\u002Fapi-soubiran-dev","---\ncontentId: f2225f4f-7da6-4515-a5bf-e6adf1c9f294\ntitle: api.soubiran.dev\ndescription: >-\n  The Laravel service that owns accounts, comments, reactions, votes, feedback,\n  and other dynamic data for my personal website.\nurl: 'https:\u002F\u002Fapi.soubiran.dev'\nrepository: 'https:\u002F\u002Fgithub.com\u002Fbarbapapazes\u002Fapi.soubiran.dev'\necosystem:\n  - type: deployment\n    name: Forge\n    description: Deploy the Laravel application and run its database migrations and metadata imports.\n    ecosystem:\n      - type: repository\n        id: api.soubiran.dev\n        name: GitHub\n        description: Host the public application source and Forge deployment script.\n        href: 'https:\u002F\u002Fgithub.com\u002Fbarbapapazes\u002Fapi.soubiran.dev'\n        ecosystem:\n          - type: stack\n            name: Laravel\n            description: Handle HTTP routes, persistence, authorization, queues, and notifications.\n            href: 'https:\u002F\u002Flaravel.com'\n          - type: stack\n            name: Laravel Sanctum\n            description: Authenticate first-party browser requests with the Laravel session.\n            href: 'https:\u002F\u002Flaravel.com\u002Fdocs\u002Fsanctum'\n          - type: stack\n            name: Laravel Socialite\n            description: Connect user accounts to GitHub, Google, and Discord.\n            href: 'https:\u002F\u002Flaravel.com\u002Fdocs\u002Fsocialite'\n          - type: stack\n            name: Filament\n            description: Provide the private administration panel for application records.\n            href: 'https:\u002F\u002Ffilamentphp.com'\n          - type: stack\n            name: Spatie Laravel Markdown\n            description: Render the Markdown preview used when writing comments.\n            href: 'https:\u002F\u002Fgithub.com\u002Fspatie\u002Flaravel-markdown'\n          - type: stack\n            name: Sentry\n            description: Capture application exceptions and performance traces.\n            href: 'https:\u002F\u002Fsentry.io'\n      - type: service\n        name: redeploy.soubiran.dev\n        description: Trigger this deployment only after an upstream website reaches production.\n  - type: data\n    name: soubiran.dev metadata\n    description: Supply stable page IDs, titles, and URLs for personal website content and infrastructure documentation.\n    ecosystem:\n      - type: website\n        name: soubiran.dev\n        description: Publish the metadata during its static build.\n  - type: website\n    relationship: consumer\n    name: soubiran.dev\n    description: Read public interaction data and send authenticated changes from the browser.\n---\n\n[api.soubiran.dev](https:\u002F\u002Fapi.soubiran.dev) owns the dynamic data behind [soubiran.dev](\u002Finfra\u002Fwebsites\u002Fsoubiran-dev). The personal website remains statically generated, while this Laravel application stores user accounts, comments, reactions, votes, feedback, messages, and notifications.\n\nIt also imports stable records for content published by [soubiran.dev](\u002Finfra\u002Fwebsites\u002Fsoubiran-dev), including its [infrastructure documentation](\u002Finfra). Stable page IDs keep those records connected when a title or URL changes.\n\n## Why the dynamic state is separate\n\nMost requests to my personal website only need static HTML and assets. Comments and reactions have different requirements. They need authenticated writes, authorization, persistence, and an administration interface. Putting those concerns in a separate application keeps them out of the static publishing path.\n\nLaravel provides the data model and request lifecycle for that mutable state. The website can still deploy as static files on Cloudflare, and its browser code calls the API only for interactive features. This split fits a content site that benefits from static delivery but still needs accounts and user-generated data.\n\nThe API stores a local page record rather than copying article content. Published records use the UUID from the source website as their API route key. Interactions point to that record, so a later content build can change the page title or canonical URL without breaking its comments or reactions.\n\n## How content and interactions meet\n\nThe website publishes `\u002Fpages.en.json` and `\u002Fpages.fr.json` catalogs. During deployment, the API fetches those catalogs and upserts their entries by UUID. The [soubiran.dev catalogs](\u002Finfra\u002Fwebsites\u002Fsoubiran-dev) supply each page's stable ID, title, description, published URL, and Markdown URL. The English catalog also includes the infrastructure documentation.\n\nThe browser addresses published-page interaction routes with that UUID. Public endpoints return emoji definitions, vote totals, feedback summaries, reactions, and comment threads. They also accept page feedback without creating an account. Comment responses include rendered Markdown, replies, like counts, author data, and capability flags, which lets the website render the thread without reproducing Laravel's authorization rules.\n\nVoting is the exception to UUID routing. A candidate article can exist in the API before publication, so vote mutations use its internal database ID. Once the article is published, a metadata import attaches the public UUID to the same record.\n\nAuthenticated routes cover user-specific data and mutations. They return the current account, votes, reactions, and notifications, and they handle comments, likes, reactions, votes, Markdown previews, and ask-me-anything messages. The API only includes private identity fields when a signed-in user reads their own account.\n\nLaravel Sanctum treats the website as a stateful first-party client. Before an unsafe request, the browser obtains a CSRF cookie and sends the request with its session credentials. Users sign in through GitHub, Google, or Discord with Laravel Socialite. This uses the browser session for the first-party website instead of exposing long-lived API tokens to client code.\n\nFilament provides the private administration panel. New accounts and user activity can create queued database notifications for administrators, while Laravel policies decide which comment and account actions each request may perform.\n\n## Deployment ordering\n\nForge deploys the application on a server. Its deployment script enables maintenance mode, pulls the selected branch, installs production PHP dependencies, optimizes Laravel, reloads PHP-FPM, and runs database migrations. It then imports the website's localized catalogs before bringing the application back online.\n\nThose imports make deployment order part of the data contract. A normal API deployment must fetch catalogs that are already live, not files from the previous website build. After a push to the website, [redeploy.soubiran.dev](\u002Finfra\u002Fservices\u002Fredeploy-soubiran-dev) waits for the corresponding Cloudflare Worker deployment to reach production before it calls the API deploy hook.\n\nThis sequence keeps static content and mutable records independent without letting their shared page IDs drift. Website builds publish the source metadata first. The API deployment imports it second. Browser requests can then attach current interaction data to the new pages.\n\n## Ecosystem\n\n::infra-ecosystem\n::\n",{"title":60,"description":170},{"loc":223},"infra\u002Fservices\u002Fapi-soubiran-dev","uoTy8_rbXZs4ep0szlOdsbwA127lHI-6FfAEfPkMda8",{"id":230,"title":231,"body":232,"contentId":787,"description":788,"ecosystem":789,"extension":17,"meta":821,"navigation":19,"part":16,"path":822,"rawbody":823,"repository":824,"robots":16,"schemaOrg":16,"seo":826,"sitemap":827,"stem":828,"url":238,"__hash__":829},"pages\u002Finfra\u002Fservices\u002Fautomation-soubiran-dev.md","automation.soubiran.dev",{"type":8,"value":233,"toc":782},[234,247,250,254,257,260,738,747,762,765,768,771,774,776,778],[64,235,236,240,241,246],{},[67,237,231],{"href":238,"rel":239},"https:\u002F\u002Fautomation.soubiran.dev",[71]," is an internal service that helps me automate various tasks with ",[67,242,245],{"href":243,"rel":244},"https:\u002F\u002Fdevelopers.cloudflare.com\u002Fworkflows\u002F",[71],"Cloudflare Workflows",".",[64,248,249],{},"There is no public access or frontend interface. The endpoint only receives calls and triggers workflows.",[90,251,253],{"id":252},"development","Development",[64,255,256],{},"The service runs on a Cloudflare Worker that receives HTTP requests from my local CLI tool. Each endpoint triggers a different Cloudflare Workflow.",[64,258,259],{},"For example, I have a workflow that I use to remind myself to publish scheduled tweets.",[261,262,267],"pre",{"className":263,"code":264,"filename":265,"language":266,"meta":11,"style":11},"language-ts shiki shiki-themes github-light github-light github-dark","import type { WorkflowEvent, WorkflowStep } from 'cloudflare:workers'\nimport { WorkflowEntrypoint } from 'cloudflare:workers'\n\nexport default {\n  async fetch(request: Request, env: Env): Promise\u003CResponse> {\n    const body = await request.json\u003C{ scheduleAt?: number, content?: string }>()\n\n    await env.AUTOMATION_WORKFLOW.create({\n      params: {\n        scheduleAt: body.scheduleAt,\n        content: body.content,\n      },\n    })\n\n    return new Response('Scheduled', { status: 201 })\n  },\n}\n\nexport class Automation extends WorkflowEntrypoint\u003CEnv, any> {\n  async run(event: WorkflowEvent\u003Cany>, step: WorkflowStep) {\n    await step.sleepUntil('trigger time', event.payload.scheduleAt)\n\n    await step.do('trigger action', async () => {\n      await fetch(this.env.DISCORD_WEBHOOK_URL, {\n        method: 'POST',\n        headers: {\n          'Content-Type': 'application\u002Fjson',\n        },\n        body: JSON.stringify({ content: event.payload.content }),\n      })\n    })\n  }\n}\n","src\u002Findex.ts","ts",[111,268,269,292,303,309,321,372,419,424,444,450,456,462,468,474,479,505,511,517,522,551,587,606,611,639,661,673,679,693,699,716,722,727,733],{"__ignoreMap":11},[270,271,274,278,281,285,288],"span",{"class":272,"line":273},"line",1,[270,275,277],{"class":276},"s8jYJ","import",[270,279,280],{"class":276}," type",[270,282,284],{"class":283},"sxrX7"," { WorkflowEvent, WorkflowStep } ",[270,286,287],{"class":276},"from",[270,289,291],{"class":290},"sIIMD"," 'cloudflare:workers'\n",[270,293,294,296,299,301],{"class":272,"line":12},[270,295,277],{"class":276},[270,297,298],{"class":283}," { WorkflowEntrypoint } ",[270,300,287],{"class":276},[270,302,291],{"class":290},[270,304,306],{"class":272,"line":305},3,[270,307,308],{"emptyLinePlaceholder":19},"\n",[270,310,312,315,318],{"class":272,"line":311},4,[270,313,314],{"class":276},"export",[270,316,317],{"class":276}," default",[270,319,320],{"class":283}," {\n",[270,322,324,327,331,334,338,341,344,347,350,352,355,358,360,363,366,369],{"class":272,"line":323},5,[270,325,326],{"class":276},"  async",[270,328,330],{"class":329},"snPdu"," fetch",[270,332,333],{"class":283},"(",[270,335,337],{"class":336},"sP4rz","request",[270,339,340],{"class":276},":",[270,342,343],{"class":329}," Request",[270,345,346],{"class":283},", ",[270,348,349],{"class":336},"env",[270,351,340],{"class":276},[270,353,354],{"class":329}," Env",[270,356,357],{"class":283},")",[270,359,340],{"class":276},[270,361,362],{"class":329}," Promise",[270,364,365],{"class":283},"\u003C",[270,367,368],{"class":329},"Response",[270,370,371],{"class":283},"> {\n",[270,373,375,378,382,385,388,391,394,397,400,403,406,408,411,413,416],{"class":272,"line":374},6,[270,376,377],{"class":276},"    const",[270,379,381],{"class":380},"sBjJW"," body",[270,383,384],{"class":276}," =",[270,386,387],{"class":276}," await",[270,389,390],{"class":283}," request.",[270,392,393],{"class":329},"json",[270,395,396],{"class":283},"\u003C{ ",[270,398,399],{"class":336},"scheduleAt",[270,401,402],{"class":276},"?:",[270,404,405],{"class":380}," number",[270,407,346],{"class":283},[270,409,410],{"class":336},"content",[270,412,402],{"class":276},[270,414,415],{"class":380}," string",[270,417,418],{"class":283}," }>()\n",[270,420,422],{"class":272,"line":421},7,[270,423,308],{"emptyLinePlaceholder":19},[270,425,427,430,433,436,438,441],{"class":272,"line":426},8,[270,428,429],{"class":276},"    await",[270,431,432],{"class":283}," env.",[270,434,435],{"class":380},"AUTOMATION_WORKFLOW",[270,437,246],{"class":283},[270,439,440],{"class":329},"create",[270,442,443],{"class":283},"({\n",[270,445,447],{"class":272,"line":446},9,[270,448,449],{"class":283},"      params: {\n",[270,451,453],{"class":272,"line":452},10,[270,454,455],{"class":283},"        scheduleAt: body.scheduleAt,\n",[270,457,459],{"class":272,"line":458},11,[270,460,461],{"class":283},"        content: body.content,\n",[270,463,465],{"class":272,"line":464},12,[270,466,467],{"class":283},"      },\n",[270,469,471],{"class":272,"line":470},13,[270,472,473],{"class":283},"    })\n",[270,475,477],{"class":272,"line":476},14,[270,478,308],{"emptyLinePlaceholder":19},[270,480,482,485,488,491,493,496,499,502],{"class":272,"line":481},15,[270,483,484],{"class":276},"    return",[270,486,487],{"class":276}," new",[270,489,490],{"class":329}," Response",[270,492,333],{"class":283},[270,494,495],{"class":290},"'Scheduled'",[270,497,498],{"class":283},", { status: ",[270,500,501],{"class":380},"201",[270,503,504],{"class":283}," })\n",[270,506,508],{"class":272,"line":507},16,[270,509,510],{"class":283},"  },\n",[270,512,514],{"class":272,"line":513},17,[270,515,516],{"class":283},"}\n",[270,518,520],{"class":272,"line":519},18,[270,521,308],{"emptyLinePlaceholder":19},[270,523,525,527,530,533,536,539,541,544,546,549],{"class":272,"line":524},19,[270,526,314],{"class":276},[270,528,529],{"class":276}," class",[270,531,532],{"class":329}," Automation",[270,534,535],{"class":276}," extends",[270,537,538],{"class":329}," WorkflowEntrypoint",[270,540,365],{"class":283},[270,542,543],{"class":329},"Env",[270,545,346],{"class":283},[270,547,548],{"class":380},"any",[270,550,371],{"class":283},[270,552,554,556,559,561,564,566,569,571,573,576,579,581,584],{"class":272,"line":553},20,[270,555,326],{"class":276},[270,557,558],{"class":329}," run",[270,560,333],{"class":283},[270,562,563],{"class":336},"event",[270,565,340],{"class":276},[270,567,568],{"class":329}," WorkflowEvent",[270,570,365],{"class":283},[270,572,548],{"class":380},[270,574,575],{"class":283},">, ",[270,577,578],{"class":336},"step",[270,580,340],{"class":276},[270,582,583],{"class":329}," WorkflowStep",[270,585,586],{"class":283},") {\n",[270,588,590,592,595,598,600,603],{"class":272,"line":589},21,[270,591,429],{"class":276},[270,593,594],{"class":283}," step.",[270,596,597],{"class":329},"sleepUntil",[270,599,333],{"class":283},[270,601,602],{"class":290},"'trigger time'",[270,604,605],{"class":283},", event.payload.scheduleAt)\n",[270,607,609],{"class":272,"line":608},22,[270,610,308],{"emptyLinePlaceholder":19},[270,612,614,616,618,621,623,626,628,631,634,637],{"class":272,"line":613},23,[270,615,429],{"class":276},[270,617,594],{"class":283},[270,619,620],{"class":329},"do",[270,622,333],{"class":283},[270,624,625],{"class":290},"'trigger action'",[270,627,346],{"class":283},[270,629,630],{"class":276},"async",[270,632,633],{"class":283}," () ",[270,635,636],{"class":276},"=>",[270,638,320],{"class":283},[270,640,642,645,647,649,652,655,658],{"class":272,"line":641},24,[270,643,644],{"class":276},"      await",[270,646,330],{"class":329},[270,648,333],{"class":283},[270,650,651],{"class":380},"this",[270,653,654],{"class":283},".env.",[270,656,657],{"class":380},"DISCORD_WEBHOOK_URL",[270,659,660],{"class":283},", {\n",[270,662,664,667,670],{"class":272,"line":663},25,[270,665,666],{"class":283},"        method: ",[270,668,669],{"class":290},"'POST'",[270,671,672],{"class":283},",\n",[270,674,676],{"class":272,"line":675},26,[270,677,678],{"class":283},"        headers: {\n",[270,680,682,685,688,691],{"class":272,"line":681},27,[270,683,684],{"class":290},"          'Content-Type'",[270,686,687],{"class":283},": ",[270,689,690],{"class":290},"'application\u002Fjson'",[270,692,672],{"class":283},[270,694,696],{"class":272,"line":695},28,[270,697,698],{"class":283},"        },\n",[270,700,702,705,708,710,713],{"class":272,"line":701},29,[270,703,704],{"class":283},"        body: ",[270,706,707],{"class":380},"JSON",[270,709,246],{"class":283},[270,711,712],{"class":329},"stringify",[270,714,715],{"class":283},"({ content: event.payload.content }),\n",[270,717,719],{"class":272,"line":718},30,[270,720,721],{"class":283},"      })\n",[270,723,725],{"class":272,"line":724},31,[270,726,473],{"class":283},[270,728,730],{"class":272,"line":729},32,[270,731,732],{"class":283},"  }\n",[270,734,736],{"class":272,"line":735},33,[270,737,516],{"class":283},[739,740,741],"blockquote",{},[64,742,743,746],{},[270,744,745],{},"!NOTE","\nThe code is simplified for clarity. The actual implementation includes error handling, logging, and other production-ready features.",[64,748,749,750,755,756,761],{},"The worker itself does not authenticate requests, even though it is exposed to the public internet. Instead, I rely on ",[67,751,754],{"href":752,"rel":753},"https:\u002F\u002Fdevelopers.cloudflare.com\u002Fcloudflare-one\u002F",[71],"Cloudflare One"," with a ",[67,757,760],{"href":758,"rel":759},"https:\u002F\u002Fdevelopers.cloudflare.com\u002Fcloudflare-one\u002Faccess-controls\u002Fservice-credentials\u002Fservice-tokens\u002F",[71],"service token"," to ensure that only my CLI tool can call the endpoints.",[64,763,764],{},"This keeps authentication outside the worker and prevents unauthorized requests from triggering workflows and generating costs.",[90,766,767],{"id":173},"Deployment",[64,769,770],{},"Cloudflare Builds deploys the service automatically. Every push to the main branch triggers a new deployment of the worker and its workflows.",[64,772,773],{},"Cloudflare's observability tools expose workflow executions and worker metrics in the dashboard.",[90,775,159],{"id":158},[161,777],{},[779,780,781],"style",{},"html pre.shiki code .s8jYJ, html code.shiki .s8jYJ{--shiki-light:#D73A49;--shiki-default:#D73A49;--shiki-dark:#F97583}html pre.shiki code .sxrX7, html code.shiki .sxrX7{--shiki-light:#24292E;--shiki-default:#24292E;--shiki-dark:#E1E4E8}html pre.shiki code .sIIMD, html code.shiki .sIIMD{--shiki-light:#032F62;--shiki-default:#032F62;--shiki-dark:#9ECBFF}html pre.shiki code .snPdu, html code.shiki .snPdu{--shiki-light:#6F42C1;--shiki-default:#6F42C1;--shiki-dark:#B392F0}html pre.shiki code .sP4rz, html code.shiki .sP4rz{--shiki-light:#E36209;--shiki-default:#E36209;--shiki-dark:#FFAB70}html pre.shiki code .sBjJW, html code.shiki .sBjJW{--shiki-light:#005CC5;--shiki-default:#005CC5;--shiki-dark:#79B8FF}html .light .shiki span {color: var(--shiki-light);background: var(--shiki-light-bg);font-style: var(--shiki-light-font-style);font-weight: var(--shiki-light-font-weight);text-decoration: var(--shiki-light-text-decoration);}html.light .shiki span {color: var(--shiki-light);background: var(--shiki-light-bg);font-style: var(--shiki-light-font-style);font-weight: var(--shiki-light-font-weight);text-decoration: var(--shiki-light-text-decoration);}html .default .shiki span {color: var(--shiki-default);background: var(--shiki-default-bg);font-style: var(--shiki-default-font-style);font-weight: var(--shiki-default-font-weight);text-decoration: var(--shiki-default-text-decoration);}html .shiki span {color: var(--shiki-default);background: var(--shiki-default-bg);font-style: var(--shiki-default-font-style);font-weight: var(--shiki-default-font-weight);text-decoration: var(--shiki-default-text-decoration);}html .dark .shiki span {color: var(--shiki-dark);background: var(--shiki-dark-bg);font-style: var(--shiki-dark-font-style);font-weight: var(--shiki-dark-font-weight);text-decoration: var(--shiki-dark-text-decoration);}html.dark .shiki span {color: var(--shiki-dark);background: var(--shiki-dark-bg);font-style: var(--shiki-dark-font-style);font-weight: var(--shiki-dark-font-weight);text-decoration: var(--shiki-dark-text-decoration);}",{"title":11,"searchDepth":12,"depth":12,"links":783},[784,785,786],{"id":252,"depth":12,"text":253},{"id":173,"depth":12,"text":767},{"id":158,"depth":12,"text":159},"640cabcb-1bd2-4960-b25b-df4f4eb2e9de","An internal service for automating tasks using Cloudflare Workflows and secured by Cloudflare One to ensure only authorized access.",[790,817],{"type":791,"name":754,"description":792,"ecosystem":793},"auth","Protect the automation endpoints so only my CLI can trigger them.",[794],{"type":173,"id":795,"name":796,"description":797,"ecosystem":798},"automation-soubiran-dev","Cloudflare Workers","Expose the automation endpoints and trigger workflows from incoming requests.",[799,814],{"type":800,"id":795,"name":801,"description":802,"ecosystem":803},"build","Cloudflare Build","Deploy the worker and workflows automatically on each push.",[804],{"type":178,"id":805,"name":179,"description":806,"ecosystem":807},"automation-repository","Host the private source code for this internal automation service.",[808,811],{"type":184,"name":809,"href":810},"Wrangler","https:\u002F\u002Fdevelopers.cloudflare.com\u002Fworkers\u002Fwrangler",{"type":184,"name":812,"href":813},"Evlog","https:\u002F\u002Fevlog.dev",{"type":815,"name":245,"description":816},"workflows","Run background jobs for the automation tasks handled by the service.",{"type":818,"name":819,"description":820},"domain","Cloudflare Domains","Manage the DNS records that route `automation.soubiran.dev` to the worker.",{},"\u002Finfra\u002Fservices\u002Fautomation-soubiran-dev","---\ncontentId: 640cabcb-1bd2-4960-b25b-df4f4eb2e9de\ntitle: automation.soubiran.dev\ndescription: >-\n  An internal service for automating tasks using Cloudflare Workflows and secured\n  by Cloudflare One to ensure only authorized access.\nurl: https:\u002F\u002Fautomation.soubiran.dev\nrepository:\n  url: https:\u002F\u002Fgithub.com\u002Fbarbapapazes\u002Fplatform\n  private: true\necosystem:\n  - type: auth\n    name: Cloudflare One\n    description: Protect the automation endpoints so only my CLI can trigger them.\n    ecosystem:\n    - type: deployment\n      id: automation-soubiran-dev\n      name: Cloudflare Workers\n      description: Expose the automation endpoints and trigger workflows from incoming requests.\n      ecosystem:\n        - type: build\n          id: automation-soubiran-dev\n          name: Cloudflare Build\n          description: Deploy the worker and workflows automatically on each push.\n          ecosystem:\n            - type: repository\n              id: automation-repository\n              name: GitHub\n              description: Host the private source code for this internal automation service.\n              ecosystem:\n                - type: stack\n                  name: Wrangler\n                  href: https:\u002F\u002Fdevelopers.cloudflare.com\u002Fworkers\u002Fwrangler\n                - type: stack\n                  name: Evlog\n                  href: https:\u002F\u002Fevlog.dev\n        - type: workflows\n          name: Cloudflare Workflows\n          description: Run background jobs for the automation tasks handled by the service.\n  - type: domain\n    name: Cloudflare Domains\n    description: Manage the DNS records that route `automation.soubiran.dev` to the worker.\n---\n\n[automation.soubiran.dev](https:\u002F\u002Fautomation.soubiran.dev) is an internal service that helps me automate various tasks with [Cloudflare Workflows](https:\u002F\u002Fdevelopers.cloudflare.com\u002Fworkflows\u002F).\n\nThere is no public access or frontend interface. The endpoint only receives calls and triggers workflows.\n\n## Development\n\nThe service runs on a Cloudflare Worker that receives HTTP requests from my local CLI tool. Each endpoint triggers a different Cloudflare Workflow.\n\nFor example, I have a workflow that I use to remind myself to publish scheduled tweets.\n\n```ts [src\u002Findex.ts]\nimport type { WorkflowEvent, WorkflowStep } from 'cloudflare:workers'\nimport { WorkflowEntrypoint } from 'cloudflare:workers'\n\nexport default {\n  async fetch(request: Request, env: Env): Promise\u003CResponse> {\n    const body = await request.json\u003C{ scheduleAt?: number, content?: string }>()\n\n    await env.AUTOMATION_WORKFLOW.create({\n      params: {\n        scheduleAt: body.scheduleAt,\n        content: body.content,\n      },\n    })\n\n    return new Response('Scheduled', { status: 201 })\n  },\n}\n\nexport class Automation extends WorkflowEntrypoint\u003CEnv, any> {\n  async run(event: WorkflowEvent\u003Cany>, step: WorkflowStep) {\n    await step.sleepUntil('trigger time', event.payload.scheduleAt)\n\n    await step.do('trigger action', async () => {\n      await fetch(this.env.DISCORD_WEBHOOK_URL, {\n        method: 'POST',\n        headers: {\n          'Content-Type': 'application\u002Fjson',\n        },\n        body: JSON.stringify({ content: event.payload.content }),\n      })\n    })\n  }\n}\n```\n\n> [!NOTE]\n> The code is simplified for clarity. The actual implementation includes error handling, logging, and other production-ready features.\n\nThe worker itself does not authenticate requests, even though it is exposed to the public internet. Instead, I rely on [Cloudflare One](https:\u002F\u002Fdevelopers.cloudflare.com\u002Fcloudflare-one\u002F) with a [service token](https:\u002F\u002Fdevelopers.cloudflare.com\u002Fcloudflare-one\u002Faccess-controls\u002Fservice-credentials\u002Fservice-tokens\u002F) to ensure that only my CLI tool can call the endpoints.\n\nThis keeps authentication outside the worker and prevents unauthorized requests from triggering workflows and generating costs.\n\n## Deployment\n\nCloudflare Builds deploys the service automatically. Every push to the main branch triggers a new deployment of the worker and its workflows.\n\nCloudflare's observability tools expose workflow executions and worker metrics in the dashboard.\n\n## Ecosystem\n\n::infra-ecosystem\n::\n",{"url":825,"private":19},"https:\u002F\u002Fgithub.com\u002Fbarbapapazes\u002Fplatform",{"title":231,"description":788},{"loc":822},"infra\u002Fservices\u002Fautomation-soubiran-dev","suOnJt7AzcGF88aUChkHxkQ87uVQEvSu_iuOczcasHM",{"id":831,"title":832,"body":833,"contentId":983,"description":984,"ecosystem":985,"extension":17,"meta":1016,"navigation":19,"part":16,"path":1017,"rawbody":1018,"repository":1019,"robots":16,"schemaOrg":16,"seo":1020,"sitemap":1021,"stem":1022,"url":839,"__hash__":1023},"pages\u002Finfra\u002Finternal-tools\u002Fcalendar-soubiran-dev.md","calendar.soubiran.dev",{"type":8,"value":834,"toc":976},[835,846,849,853,856,859,862,866,885,902,913,917,935,946,949,953,969,972,974],[64,836,837,841,842,845],{},[67,838,832],{"href":839,"rel":840},"https:\u002F\u002Fcalendar.soubiran.dev",[71]," publishes my LinkedIn content schedule as a private iCalendar feed. A command in my content-creation CLI turns dated Markdown files into calendar events, uploads one ",[111,843,844],{},".ics"," file to Cloudflare R2, and gives my calendar client a tokenized subscription URL.",[64,847,848],{},"The hostname is a delivery endpoint rather than a calendar interface. The Worker reads the generated file and returns it to authorized calendar clients. This keeps the private content repository outside the request path.",[90,850,852],{"id":851},"why-i-publish-a-complete-calendar-file","Why I publish a complete calendar file",[64,854,855],{},"The schedule is a derived view of content I already manage as Markdown. Generating it during my publishing workflow means the Markdown remains the source of truth, while calendar applications receive a format they understand without gaining access to the repository.",[64,857,858],{},"I store the complete feed as one R2 object instead of rebuilding it on every request. The Worker does not scan directories, parse frontmatter, or assemble events. It only checks access and reads the current object. Replacing that object updates every subscriber without a Worker deployment.",[64,860,861],{},"This split fits internal feeds derived from files or another authoring system. Generation can use trusted local context, while delivery stays small and read-only.",[90,863,865],{"id":864},"how-markdown-becomes-calendar-events","How Markdown becomes calendar events",[64,867,868,869,872,873,876,877,880,881,884],{},"The ",[111,870,871],{},"linkedin publish-calendar"," workflow runs through the content-creation CLI in the private ",[111,874,875],{},"platform"," monorepo. It scans the selected content directory for dated ",[111,878,879],{},"YYYY\u002FMM\u002FDD"," folders and reads each ",[111,882,883],{},"linkedin.md"," file. Frontmatter provides the title, theme, publication state, and planned media.",[64,886,887,888,346,891,346,894,897,898,901],{},"The generator sorts the entries by date and creates one all-day event for each file. Its summary combines the content title with a status such as ",[111,889,890],{},"Not ready",[111,892,893],{},"To be recorded",[111,895,896],{},"To be published",", or ",[111,899,900],{},"Published",". The description links back to the private source file and its local VS Code path, with theme and media details when present.",[64,903,904,905,908,909,912],{},"After serializing the events as iCalendar data, the CLI uploads the result to the ",[111,906,907],{},"content-creation"," R2 bucket under the fixed ",[111,910,911],{},"content-creation.ics"," key. It then prints the subscription URL with the configured calendar token. The calendar artifact changes independently of the Worker code because publication replaces the object directly.",[90,914,916],{"id":915},"how-private-delivery-works","How private delivery works",[64,918,919,920,114,923,926,927,930,931,934],{},"The Calendar Worker exposes ",[111,921,922],{},"GET",[111,924,925],{},"HEAD"," requests for ",[111,928,929],{},"\u002Fcontent-creation.ics",". It compares the ",[111,932,933],{},"token"," query parameter with a Worker secret through a timing-safe equality check, then reads the fixed object through its R2 binding. Invalid credentials never reach storage.",[64,936,937,938,941,942,945],{},"The response uses the ",[111,939,940],{},"text\u002Fcalendar"," content type and disables shared caching with ",[111,943,944],{},"private, no-store",". This boundary matters because event descriptions contain links into my private content repository and local editor paths. The token is a subscription credential, not part of the calendar data, and the Worker does not write it to structured logs.",[64,947,948],{},"Evlog records the requested route, whether a credential was present, the R2 object metadata, and the response result. These events show delivery failures without exposing the feed or its token.",[90,950,952],{"id":951},"how-it-runs","How it runs",[64,954,955,956,962,963,965,966,968],{},"The Calendar Worker and the content-creation CLI live in the private ",[67,957,959,961],{"href":825,"rel":958},[71],[111,960,875],{}," repository",". The Worker's Wrangler configuration maps ",[111,964,832],{}," to the deployment, binds the ",[111,967,907],{}," R2 bucket, and enables Cloudflare observability. Its deployment command publishes the Worker with Wrangler.",[64,970,971],{},"Calendar publication follows a separate path. I run the CLI against the content repository when its LinkedIn schedule changes, and Wrangler uploads the generated object to remote R2. Existing calendar subscriptions read that new object on their next request, so a content update does not rebuild or redeploy the delivery service.",[90,973,159],{"id":158},[161,975],{},{"title":11,"searchDepth":12,"depth":12,"links":977},[978,979,980,981,982],{"id":851,"depth":12,"text":852},{"id":864,"depth":12,"text":865},{"id":915,"depth":12,"text":916},{"id":951,"depth":12,"text":952},{"id":158,"depth":12,"text":159},"31d1ccc9-a1e5-4fb5-b84a-736b91d8f360","The private calendar feed that turns scheduled LinkedIn Markdown into iCalendar events and serves them from R2 through a token-protected Worker.",[986,1014],{"type":173,"id":987,"name":796,"description":988,"ecosystem":989},"calendar-soubiran-dev","Authorize calendar subscriptions and return the current iCalendar object.",[990,998],{"type":178,"id":987,"name":179,"description":991,"href":825,"ecosystem":992},"Host the private Worker and content-creation CLI source.",[993,996],{"type":184,"name":809,"description":994,"href":995},"Configure the custom domain and R2 binding, then deploy the Worker.","https:\u002F\u002Fdevelopers.cloudflare.com\u002Fworkers\u002Fwrangler\u002F",{"type":184,"name":812,"description":997,"href":813},"Record structured delivery events without recording the calendar token.",{"type":999,"name":1000,"description":1001,"ecosystem":1002},"object-storage","Cloudflare R2","Store the generated calendar under the fixed `content-creation.ics` key.",[1003],{"type":212,"name":1004,"description":1005,"ecosystem":1006},"LinkedIn iCalendar feed","Hold the complete set of dated LinkedIn entries as all-day events.",[1007],{"type":800,"name":1008,"description":1009,"ecosystem":1010},"Content creation CLI","Generate the complete feed and upload it to R2.",[1011],{"type":212,"name":1012,"description":1013},"LinkedIn Markdown","Supply dated publication state and metadata from the private content repository.",{"type":818,"name":819,"description":1015},"Route `calendar.soubiran.dev` to the Worker.",{},"\u002Finfra\u002Finternal-tools\u002Fcalendar-soubiran-dev","---\ncontentId: 31d1ccc9-a1e5-4fb5-b84a-736b91d8f360\ntitle: calendar.soubiran.dev\ndescription: >-\n  The private calendar feed that turns scheduled LinkedIn Markdown into\n  iCalendar events and serves them from R2 through a token-protected Worker.\nurl: 'https:\u002F\u002Fcalendar.soubiran.dev'\nrepository:\n  url: 'https:\u002F\u002Fgithub.com\u002Fbarbapapazes\u002Fplatform'\n  private: true\necosystem:\n  - type: deployment\n    id: calendar-soubiran-dev\n    name: Cloudflare Workers\n    description: Authorize calendar subscriptions and return the current iCalendar object.\n    ecosystem:\n      - type: repository\n        id: calendar-soubiran-dev\n        name: GitHub\n        description: Host the private Worker and content-creation CLI source.\n        href: 'https:\u002F\u002Fgithub.com\u002Fbarbapapazes\u002Fplatform'\n        ecosystem:\n          - type: stack\n            name: Wrangler\n            description: Configure the custom domain and R2 binding, then deploy the Worker.\n            href: 'https:\u002F\u002Fdevelopers.cloudflare.com\u002Fworkers\u002Fwrangler\u002F'\n          - type: stack\n            name: Evlog\n            description: Record structured delivery events without recording the calendar token.\n            href: 'https:\u002F\u002Fevlog.dev'\n      - type: object-storage\n        name: Cloudflare R2\n        description: Store the generated calendar under the fixed `content-creation.ics` key.\n        ecosystem:\n          - type: data\n            name: LinkedIn iCalendar feed\n            description: Hold the complete set of dated LinkedIn entries as all-day events.\n            ecosystem:\n              - type: build\n                name: Content creation CLI\n                description: Generate the complete feed and upload it to R2.\n                ecosystem:\n                  - type: data\n                    name: LinkedIn Markdown\n                    description: Supply dated publication state and metadata from the private content repository.\n  - type: domain\n    name: Cloudflare Domains\n    description: Route `calendar.soubiran.dev` to the Worker.\n---\n\n[calendar.soubiran.dev](https:\u002F\u002Fcalendar.soubiran.dev) publishes my LinkedIn content schedule as a private iCalendar feed. A command in my content-creation CLI turns dated Markdown files into calendar events, uploads one `.ics` file to Cloudflare R2, and gives my calendar client a tokenized subscription URL.\n\nThe hostname is a delivery endpoint rather than a calendar interface. The Worker reads the generated file and returns it to authorized calendar clients. This keeps the private content repository outside the request path.\n\n## Why I publish a complete calendar file\n\nThe schedule is a derived view of content I already manage as Markdown. Generating it during my publishing workflow means the Markdown remains the source of truth, while calendar applications receive a format they understand without gaining access to the repository.\n\nI store the complete feed as one R2 object instead of rebuilding it on every request. The Worker does not scan directories, parse frontmatter, or assemble events. It only checks access and reads the current object. Replacing that object updates every subscriber without a Worker deployment.\n\nThis split fits internal feeds derived from files or another authoring system. Generation can use trusted local context, while delivery stays small and read-only.\n\n## How Markdown becomes calendar events\n\nThe `linkedin publish-calendar` workflow runs through the content-creation CLI in the private `platform` monorepo. It scans the selected content directory for dated `YYYY\u002FMM\u002FDD` folders and reads each `linkedin.md` file. Frontmatter provides the title, theme, publication state, and planned media.\n\nThe generator sorts the entries by date and creates one all-day event for each file. Its summary combines the content title with a status such as `Not ready`, `To be recorded`, `To be published`, or `Published`. The description links back to the private source file and its local VS Code path, with theme and media details when present.\n\nAfter serializing the events as iCalendar data, the CLI uploads the result to the `content-creation` R2 bucket under the fixed `content-creation.ics` key. It then prints the subscription URL with the configured calendar token. The calendar artifact changes independently of the Worker code because publication replaces the object directly.\n\n## How private delivery works\n\nThe Calendar Worker exposes `GET` and `HEAD` requests for `\u002Fcontent-creation.ics`. It compares the `token` query parameter with a Worker secret through a timing-safe equality check, then reads the fixed object through its R2 binding. Invalid credentials never reach storage.\n\nThe response uses the `text\u002Fcalendar` content type and disables shared caching with `private, no-store`. This boundary matters because event descriptions contain links into my private content repository and local editor paths. The token is a subscription credential, not part of the calendar data, and the Worker does not write it to structured logs.\n\nEvlog records the requested route, whether a credential was present, the R2 object metadata, and the response result. These events show delivery failures without exposing the feed or its token.\n\n## How it runs\n\nThe Calendar Worker and the content-creation CLI live in the private [`platform` repository](https:\u002F\u002Fgithub.com\u002Fbarbapapazes\u002Fplatform). The Worker's Wrangler configuration maps `calendar.soubiran.dev` to the deployment, binds the `content-creation` R2 bucket, and enables Cloudflare observability. Its deployment command publishes the Worker with Wrangler.\n\nCalendar publication follows a separate path. I run the CLI against the content repository when its LinkedIn schedule changes, and Wrangler uploads the generated object to remote R2. Existing calendar subscriptions read that new object on their next request, so a content update does not rebuild or redeploy the delivery service.\n\n## Ecosystem\n\n::infra-ecosystem\n::\n",{"url":825,"private":19},{"title":832,"description":984},{"loc":1017},"infra\u002Finternal-tools\u002Fcalendar-soubiran-dev","KUhVjQNZ6QirsS_Rih2rZnq0YTPAy6l9nNU3chHoafE",{"id":1025,"title":1026,"body":1027,"contentId":1128,"description":1129,"ecosystem":1130,"extension":17,"meta":1177,"navigation":19,"part":16,"path":1178,"rawbody":1179,"repository":1107,"robots":16,"schemaOrg":16,"seo":1180,"sitemap":1181,"stem":1182,"url":1033,"__hash__":1183},"pages\u002Finfra\u002Fwebsites\u002Fcode-soubiran-dev.md","code.soubiran.dev",{"type":8,"value":1028,"toc":1121},[1029,1036,1043,1047,1050,1053,1056,1060,1063,1066,1073,1077,1088,1094,1097,1100,1102,1114,1117,1119],[64,1030,1031,1035],{},[67,1032,1026],{"href":1033,"rel":1034},"https:\u002F\u002Fcode.soubiran.dev",[71]," turns source code into syntax-highlighted PNG images. It combines a browser editor for people with a public MCP endpoint for assistants, and both paths use the same page to render the final image.",[64,1037,1038,1039,1042],{},"The browser exposes URL-backed controls for the code, language, canvas size, gradient, title, and watermark before downloading the result. An MCP client can send those fields to ",[111,1040,1041],{},"https:\u002F\u002Fcode.soubiran.dev\u002Fmcp"," and receive the PNG as image content.",[90,1044,1046],{"id":1045},"why-the-url-holds-the-editor-state","Why the URL holds the editor state",[64,1048,1049],{},"I wanted shared links and automated captures to reproduce the same image without storing documents on a server. The app therefore keeps its editor state in URL search parameters. It encodes the source code as Base64 over UTF-8 bytes and stores the other options as plain parameter values.",[64,1051,1052],{},"This URL is the contract between the two ways to use the product. Opening it reconstructs the editor in a browser. The MCP Worker can build the same URL from tool arguments and pass it to a remote browser without maintaining a second rendering implementation.",[64,1054,1055],{},"The approach fits a small rendering tool whose state can travel with the link and does not need persistence. Base64 turns the UTF-8 source bytes into text that the query parameter can carry. It does not encrypt them, so these URLs are not intended for secrets.",[90,1057,1059],{"id":1058},"how-the-browser-creates-an-image","How the browser creates an image",[64,1061,1062],{},"The Vue app reads the URL when it starts. Each URL-backed control writes its new value back to the query string, while empty and default values stay out of the URL where possible. Copying the current address is enough to preserve the image configuration.",[64,1064,1065],{},"The editor renders two aligned layers. Shiki produces highlighted HTML underneath a transparent textarea, so users edit normal text while seeing highlighted code. A wrapper around the editor, title, background, and watermark marks the exact DOM element used for capture.",[64,1067,1068,1069,1072],{},"When the user selects Capture, ",[111,1070,1071],{},"modern-screenshot"," converts that element into a PNG data URL and the browser downloads it. This capture path does not call the Worker or an image-generation API.",[90,1074,1076],{"id":1075},"how-the-mcp-route-reuses-the-page","How the MCP route reuses the page",[64,1078,1079,1080,1083,1084,1087],{},"The Cloudflare Worker handles ",[111,1081,1082],{},"\u002Fmcp"," and exposes one tool, ",[111,1085,1086],{},"generate_code_image",". Its input uses the URL-backed image fields: code, language, size, gradient, title, and watermark.",[64,1089,1090,1091,1093],{},"For each call, the Worker validates the arguments and creates a ",[111,1092,1026],{}," URL. It omits options that match the defaults and uses the same Base64 encoder as the browser for source code. The Worker then asks Cloudflare Browser Run to open that public URL, wait for the marked editor element, and capture it as a PNG.",[64,1095,1096],{},"Browser Run returns the image bytes to the Worker. The MCP response carries those bytes as Base64-encoded image content with its MIME type. Reusing the public page keeps browser downloads and MCP results on the same Vue, Shiki, and CSS rendering path.",[64,1098,1099],{},"The MCP endpoint is public, but its Browser Run credentials remain in Worker environment bindings. Tool logs record option choices, source length, outcome, duration, and image metadata without recording the source code, title, or watermark.",[90,1101,952],{"id":951},[64,1103,1104,1105,1110,1111,1113],{},"Vite builds the Vue app and the Worker from the ",[67,1106,1109],{"href":1107,"rel":1108},"https:\u002F\u002Fgithub.com\u002Fbarbapapazes\u002Fcode.soubiran.dev",[71],"same repository",". Cloudflare Workers Static Assets serves the app, while the asset configuration sends ",[111,1112,1082],{}," to the Worker before applying the single-page application fallback. The editor and its automation endpoint therefore share one deployment and one domain.",[64,1115,1116],{},"The deployment command builds the project before Wrangler publishes it. GitHub Actions checks pushes and pull requests with linting, a production build, type checking, and tests. Cloudflare logs and traces cover Worker requests, including MCP tool outcomes.",[90,1118,159],{"id":158},[161,1120],{},{"title":11,"searchDepth":12,"depth":12,"links":1122},[1123,1124,1125,1126,1127],{"id":1045,"depth":12,"text":1046},{"id":1058,"depth":12,"text":1059},{"id":1075,"depth":12,"text":1076},{"id":951,"depth":12,"text":952},{"id":158,"depth":12,"text":159},"2e9e7543-2215-49a6-b37d-a531f779b2da","The browser editor and public MCP endpoint that turn URL-based editor state into syntax-highlighted PNG images.",[1131,1175],{"type":173,"id":1132,"name":796,"description":1133,"ecosystem":1134},"code-soubiran-dev","Serve the browser app and handle the public `\u002Fmcp` endpoint.",[1135,1171],{"type":178,"id":1026,"name":179,"description":1136,"href":1107,"ecosystem":1137},"Host the editor, Worker, shared image contract, and deployment configuration.",[1138,1142,1146,1150,1154,1157,1161,1165,1167],{"type":184,"name":1139,"description":1140,"href":1141},"Vite","Build the browser app and Worker from one project.","https:\u002F\u002Fvite.dev",{"type":184,"name":1143,"description":1144,"href":1145},"Vue","Bind the editor controls to URL search parameters.","https:\u002F\u002Fvuejs.org",{"type":184,"name":1147,"description":1148,"href":1149},"Nuxt UI","Provide the editor controls and visual components.","https:\u002F\u002Fui.nuxt.com",{"type":184,"name":1151,"description":1152,"href":1153},"Shiki","Render the syntax-highlighted code layer in the browser.","https:\u002F\u002Fshiki.style",{"type":184,"name":1071,"description":1155,"href":1156},"Capture the rendered editor element as a PNG in the browser.","https:\u002F\u002Fgithub.com\u002Fqq15725\u002Fmodern-screenshot",{"type":184,"name":1158,"description":1159,"href":1160},"Model Context Protocol SDK","Define the image-generation tool and its image response.","https:\u002F\u002Fmodelcontextprotocol.io",{"type":184,"name":1162,"description":1163,"href":1164},"Cloudflare Agents","Expose the MCP server through the Worker's `\u002Fmcp` route.","https:\u002F\u002Fdevelopers.cloudflare.com\u002Fagents\u002F",{"type":184,"name":812,"description":1166,"href":813},"Record structured MCP tool telemetry without logging image content.",{"type":1168,"relationship":220,"name":1169,"description":1170},"ci\u002Fcd","GitHub Actions","Lint, build, type-check, and test changes pushed to the repository.",{"type":184,"name":1172,"description":1173,"href":1174},"Cloudflare Browser Run","Load a URL that describes an image and capture its rendered editor element.","https:\u002F\u002Fdevelopers.cloudflare.com\u002Fbrowser-rendering\u002F",{"type":818,"name":819,"description":1176},"Route `code.soubiran.dev` to the Worker deployment.",{},"\u002Finfra\u002Fwebsites\u002Fcode-soubiran-dev","---\ncontentId: 2e9e7543-2215-49a6-b37d-a531f779b2da\ntitle: code.soubiran.dev\ndescription: >-\n  The browser editor and public MCP endpoint that turn URL-based editor state\n  into syntax-highlighted PNG images.\nurl: 'https:\u002F\u002Fcode.soubiran.dev'\nrepository: 'https:\u002F\u002Fgithub.com\u002Fbarbapapazes\u002Fcode.soubiran.dev'\necosystem:\n  - type: deployment\n    id: code-soubiran-dev\n    name: Cloudflare Workers\n    description: Serve the browser app and handle the public `\u002Fmcp` endpoint.\n    ecosystem:\n      - type: repository\n        id: code.soubiran.dev\n        name: GitHub\n        description: Host the editor, Worker, shared image contract, and deployment configuration.\n        href: 'https:\u002F\u002Fgithub.com\u002Fbarbapapazes\u002Fcode.soubiran.dev'\n        ecosystem:\n          - type: stack\n            name: Vite\n            description: Build the browser app and Worker from one project.\n            href: 'https:\u002F\u002Fvite.dev'\n          - type: stack\n            name: Vue\n            description: Bind the editor controls to URL search parameters.\n            href: 'https:\u002F\u002Fvuejs.org'\n          - type: stack\n            name: Nuxt UI\n            description: Provide the editor controls and visual components.\n            href: 'https:\u002F\u002Fui.nuxt.com'\n          - type: stack\n            name: Shiki\n            description: Render the syntax-highlighted code layer in the browser.\n            href: 'https:\u002F\u002Fshiki.style'\n          - type: stack\n            name: modern-screenshot\n            description: Capture the rendered editor element as a PNG in the browser.\n            href: 'https:\u002F\u002Fgithub.com\u002Fqq15725\u002Fmodern-screenshot'\n          - type: stack\n            name: Model Context Protocol SDK\n            description: Define the image-generation tool and its image response.\n            href: 'https:\u002F\u002Fmodelcontextprotocol.io'\n          - type: stack\n            name: Cloudflare Agents\n            description: Expose the MCP server through the Worker's `\u002Fmcp` route.\n            href: 'https:\u002F\u002Fdevelopers.cloudflare.com\u002Fagents\u002F'\n          - type: stack\n            name: Evlog\n            description: Record structured MCP tool telemetry without logging image content.\n            href: 'https:\u002F\u002Fevlog.dev'\n          - type: ci\u002Fcd\n            relationship: consumer\n            name: GitHub Actions\n            description: Lint, build, type-check, and test changes pushed to the repository.\n      - type: stack\n        name: Cloudflare Browser Run\n        description: Load a URL that describes an image and capture its rendered editor element.\n        href: 'https:\u002F\u002Fdevelopers.cloudflare.com\u002Fbrowser-rendering\u002F'\n  - type: domain\n    name: Cloudflare Domains\n    description: Route `code.soubiran.dev` to the Worker deployment.\n---\n\n[code.soubiran.dev](https:\u002F\u002Fcode.soubiran.dev) turns source code into syntax-highlighted PNG images. It combines a browser editor for people with a public MCP endpoint for assistants, and both paths use the same page to render the final image.\n\nThe browser exposes URL-backed controls for the code, language, canvas size, gradient, title, and watermark before downloading the result. An MCP client can send those fields to `https:\u002F\u002Fcode.soubiran.dev\u002Fmcp` and receive the PNG as image content.\n\n## Why the URL holds the editor state\n\nI wanted shared links and automated captures to reproduce the same image without storing documents on a server. The app therefore keeps its editor state in URL search parameters. It encodes the source code as Base64 over UTF-8 bytes and stores the other options as plain parameter values.\n\nThis URL is the contract between the two ways to use the product. Opening it reconstructs the editor in a browser. The MCP Worker can build the same URL from tool arguments and pass it to a remote browser without maintaining a second rendering implementation.\n\nThe approach fits a small rendering tool whose state can travel with the link and does not need persistence. Base64 turns the UTF-8 source bytes into text that the query parameter can carry. It does not encrypt them, so these URLs are not intended for secrets.\n\n## How the browser creates an image\n\nThe Vue app reads the URL when it starts. Each URL-backed control writes its new value back to the query string, while empty and default values stay out of the URL where possible. Copying the current address is enough to preserve the image configuration.\n\nThe editor renders two aligned layers. Shiki produces highlighted HTML underneath a transparent textarea, so users edit normal text while seeing highlighted code. A wrapper around the editor, title, background, and watermark marks the exact DOM element used for capture.\n\nWhen the user selects Capture, `modern-screenshot` converts that element into a PNG data URL and the browser downloads it. This capture path does not call the Worker or an image-generation API.\n\n## How the MCP route reuses the page\n\nThe Cloudflare Worker handles `\u002Fmcp` and exposes one tool, `generate_code_image`. Its input uses the URL-backed image fields: code, language, size, gradient, title, and watermark.\n\nFor each call, the Worker validates the arguments and creates a `code.soubiran.dev` URL. It omits options that match the defaults and uses the same Base64 encoder as the browser for source code. The Worker then asks Cloudflare Browser Run to open that public URL, wait for the marked editor element, and capture it as a PNG.\n\nBrowser Run returns the image bytes to the Worker. The MCP response carries those bytes as Base64-encoded image content with its MIME type. Reusing the public page keeps browser downloads and MCP results on the same Vue, Shiki, and CSS rendering path.\n\nThe MCP endpoint is public, but its Browser Run credentials remain in Worker environment bindings. Tool logs record option choices, source length, outcome, duration, and image metadata without recording the source code, title, or watermark.\n\n## How it runs\n\nVite builds the Vue app and the Worker from the [same repository](https:\u002F\u002Fgithub.com\u002Fbarbapapazes\u002Fcode.soubiran.dev). Cloudflare Workers Static Assets serves the app, while the asset configuration sends `\u002Fmcp` to the Worker before applying the single-page application fallback. The editor and its automation endpoint therefore share one deployment and one domain.\n\nThe deployment command builds the project before Wrangler publishes it. GitHub Actions checks pushes and pull requests with linting, a production build, type checking, and tests. Cloudflare logs and traces cover Worker requests, including MCP tool outcomes.\n\n## Ecosystem\n\n::infra-ecosystem\n::\n",{"title":1026,"description":1129},{"loc":1178},"infra\u002Fwebsites\u002Fcode-soubiran-dev","k7DAtWtOXlBf89HINRnZzBBEM062q8JSV1rdXVwG7yw",{"id":1185,"title":1186,"body":1187,"contentId":1523,"description":1524,"ecosystem":1525,"extension":17,"meta":1573,"navigation":19,"part":16,"path":1574,"rawbody":1575,"repository":1200,"robots":16,"schemaOrg":16,"seo":1576,"sitemap":1577,"stem":1578,"url":1193,"__hash__":1579},"pages\u002Finfra\u002Fservices\u002Fmcp-soubiran-dev.md","mcp.soubiran.dev",{"type":8,"value":1188,"toc":1512},[1189,1196,1210,1214,1233,1240,1248,1251,1255,1261,1271,1274,1285,1289,1295,1298,1302,1308,1311,1457,1463,1467,1472,1492,1495,1499,1502,1505,1507,1509],[64,1190,1191,1195],{},[67,1192,1186],{"href":1193,"rel":1194},"https:\u002F\u002Fmcp.soubiran.dev\u002Fmcp",[71]," gives assistants access to my public English and French content. It exposes three read-only tools for topic discovery, complete Markdown retrieval, and exact metadata queries.",[64,1197,868,1198,1203,1204,1206,1207,1209],{},[67,1199,1202],{"href":1200,"rel":1201},"https:\u002F\u002Fgithub.com\u002Fbarbapapazes\u002Fmcp.soubiran.dev",[71],"TypeScript Worker"," uses the Model Context Protocol SDK and the Cloudflare Agents MCP handler. Requests to ",[111,1205,1082],{}," reach the server. Other paths redirect to ",[67,1208,76],{"href":75},". The handler does not require authentication.",[90,1211,1213],{"id":1212},"one-content-catalog-two-locales","One content catalog, two locales",[64,1215,1216,1217,114,1219,1221,1222,114,1225,1228,1229,1232],{},"The server loads the website's ",[111,1218,113],{},[111,1220,117],{}," feeds in parallel. ",[111,1223,1224],{},"PAGES_EN_URL",[111,1226,1227],{},"PAGES_FR_URL"," configure their locations. Zod validates the records, and the loader combines them into one ",[111,1230,1231],{},"pages"," array.",[64,1234,1235,1236,1239],{},"Each record has a stable ID, locale, content type, title, published URL, and ",[111,1237,1238],{},"markdownUrl",". Descriptions, dates, and translation links are optional. Relative URLs are resolved against the feed URL. Duplicate IDs across the combined feeds cause an error rather than selecting one document arbitrarily.",[64,1241,1242,1243,1247],{},"The website catalog covers pages, posts, talks, series, series articles, and infrastructure documentation. ",[67,1244,1246],{"href":1245},"\u002Finfra\u002Fwebsites\u002Ftalks-soubiran-dev","talks.soubiran.dev"," supplies talk content to the website during its build. The MCP reads the resulting website documents through the same catalog as the rest of the content.",[64,1249,1250],{},"Publication stays with the website. The MCP fetches metadata and follows the supplied Markdown URLs; it does not maintain another document store.",[90,1252,1254],{"id":1253},"search-retrieve-or-query","Search, retrieve, or query",[1256,1257,1259],"h3",{"id":1258},"search_content",[111,1260,1258],{},[64,1262,1263,1264,1267,1268,246],{},"This tool accepts a natural-language query and searches the ",[111,1265,1266],{},"soubiran-dev"," Cloudflare AI Search instance. Search and catalog loading run in parallel. Retrieval requests up to 20 results with a match threshold of 0.6 and reranks them with ",[111,1269,1270],{},"@cf\u002Fbaai\u002Fbge-reranker-base",[64,1272,1273],{},"The server matches each search chunk's metadata ID to a catalog record. It returns the score, excerpt, source key, and page metadata, including the published URL for citations. Chunks without a matching catalog entry are omitted.",[64,1275,1276,1277,1280,1281,1284],{},"Search finds relevant content, but ranking does not establish which page is newest. Exact date comparisons belong in ",[111,1278,1279],{},"list_pages",", and an assistant should read the full document with ",[111,1282,1283],{},"get_page"," before summarizing it.",[1256,1286,1287],{"id":1283},[111,1288,1283],{},[64,1290,1291,1292,1294],{},"This tool accepts an exact stable ID, resolves it in the combined catalog, and fetches its ",[111,1293,1238],{},". It returns the complete document as text. An unknown ID or failed upstream request produces an MCP tool error.",[64,1296,1297],{},"A talk document can include its summary, event details, resource links, and transcript. The website composes that Markdown, so retrieval works the same way for a talk, post, or infrastructure page.",[1256,1299,1300],{"id":1279},[111,1301,1279],{},[64,1303,1304,1305,1307],{},"This tool accepts a single asynchronous JavaScript arrow function. The function can read the combined ",[111,1306,1231],{}," array to filter by locale or type, match titles, compare dates, count records, or group metadata. It does not search document bodies.",[64,1309,1310],{},"For example, this query returns the five most recent dated English posts with the fields needed for retrieval and citation:",[261,1312,1316],{"className":1313,"code":1314,"language":1315,"meta":11,"style":11},"language-js shiki shiki-themes github-light github-light github-dark","async () => pages\n  .filter(page => page.locale === 'en' && page.type === 'post' && page.date)\n  .sort((a, b) => b.date.localeCompare(a.date))\n  .slice(0, 5)\n  .map(({ id, title, date, url }) => ({ id, title, date, url }))\n","js",[111,1317,1318,1329,1370,1401,1421],{"__ignoreMap":11},[270,1319,1320,1322,1324,1326],{"class":272,"line":273},[270,1321,630],{"class":276},[270,1323,633],{"class":283},[270,1325,636],{"class":276},[270,1327,1328],{"class":283}," pages\n",[270,1330,1331,1334,1337,1339,1342,1345,1348,1351,1354,1357,1360,1362,1365,1367],{"class":272,"line":12},[270,1332,1333],{"class":283},"  .",[270,1335,1336],{"class":329},"filter",[270,1338,333],{"class":283},[270,1340,1341],{"class":336},"page",[270,1343,1344],{"class":276}," =>",[270,1346,1347],{"class":283}," page.locale ",[270,1349,1350],{"class":276},"===",[270,1352,1353],{"class":290}," 'en'",[270,1355,1356],{"class":276}," &&",[270,1358,1359],{"class":283}," page.type ",[270,1361,1350],{"class":276},[270,1363,1364],{"class":290}," 'post'",[270,1366,1356],{"class":276},[270,1368,1369],{"class":283}," page.date)\n",[270,1371,1372,1374,1377,1380,1382,1384,1387,1390,1392,1395,1398],{"class":272,"line":305},[270,1373,1333],{"class":283},[270,1375,1376],{"class":329},"sort",[270,1378,1379],{"class":283},"((",[270,1381,67],{"class":336},[270,1383,346],{"class":283},[270,1385,1386],{"class":336},"b",[270,1388,1389],{"class":283},") ",[270,1391,636],{"class":276},[270,1393,1394],{"class":283}," b.date.",[270,1396,1397],{"class":329},"localeCompare",[270,1399,1400],{"class":283},"(a.date))\n",[270,1402,1403,1405,1408,1410,1413,1415,1418],{"class":272,"line":311},[270,1404,1333],{"class":283},[270,1406,1407],{"class":329},"slice",[270,1409,333],{"class":283},[270,1411,1412],{"class":380},"0",[270,1414,346],{"class":283},[270,1416,1417],{"class":380},"5",[270,1419,1420],{"class":283},")\n",[270,1422,1423,1425,1428,1431,1434,1436,1439,1441,1444,1446,1449,1452,1454],{"class":272,"line":323},[270,1424,1333],{"class":283},[270,1426,1427],{"class":329},"map",[270,1429,1430],{"class":283},"(({ ",[270,1432,1433],{"class":336},"id",[270,1435,346],{"class":283},[270,1437,1438],{"class":336},"title",[270,1440,346],{"class":283},[270,1442,1443],{"class":336},"date",[270,1445,346],{"class":283},[270,1447,1448],{"class":336},"url",[270,1450,1451],{"class":283}," }) ",[270,1453,636],{"class":276},[270,1455,1456],{"class":283}," ({ id, title, date, url }))\n",[64,1458,1459,1460,1462],{},"Returning only the relevant records and fields reduces how much metadata enters the model's context. Both discovery tools return IDs that ",[111,1461,1283],{}," can resolve.",[90,1464,1466],{"id":1465},"isolated-code-execution","Isolated code execution",[64,1468,1469,1471],{},[111,1470,1279],{}," uses Code Mode to answer metadata questions without a separate tool for every filter or aggregation. Acorn parses the submitted code and checks that it is one asynchronous arrow function. Input is limited to 20,000 characters.",[64,1473,1474,1475,1477,1478,1481,1482,346,1485,1488,1489,246],{},"Worker Loader runs the function in a temporary Worker with a recursively frozen copy of ",[111,1476,1231],{},". The Worker has no secrets, storage, or application bindings. Network access is disabled with ",[111,1479,1480],{},"globalOutbound: null",", and the executor also disables ",[111,1483,1484],{},"fetch",[111,1486,1487],{},"WebSocket",", and ",[111,1490,1491],{},"EventSource",[64,1493,1494],{},"Execution has a 50 ms CPU limit and permits no subrequests. The function must return a JSON-serializable value, and the server rejects results larger than 256 KiB. Query failures become tool errors rather than failing the MCP request handler.",[90,1496,1498],{"id":1497},"deployment-and-observability","Deployment and observability",[64,1500,1501],{},"Wrangler deploys the server to Cloudflare Workers with an AI Search namespace binding and a Worker Loader binding. The MCP handler uses a configurable allowed CORS origin.",[64,1503,1504],{},"Cloudflare persists logs and traces. Sentry captures exceptions, with user information and HTTP body collection disabled. Evlog records request duration, tool outcomes, selected upstream timings, result counts, and response sizes. Submitted query code is not logged. Redaction rules cover input and output payloads, queries, document content, excerpts, Markdown, authorization headers, and cookies.",[90,1506,159],{"id":158},[161,1508],{},[779,1510,1511],{},"html pre.shiki code .s8jYJ, html code.shiki .s8jYJ{--shiki-light:#D73A49;--shiki-default:#D73A49;--shiki-dark:#F97583}html pre.shiki code .sxrX7, html code.shiki .sxrX7{--shiki-light:#24292E;--shiki-default:#24292E;--shiki-dark:#E1E4E8}html pre.shiki code .snPdu, html code.shiki .snPdu{--shiki-light:#6F42C1;--shiki-default:#6F42C1;--shiki-dark:#B392F0}html pre.shiki code .sP4rz, html code.shiki .sP4rz{--shiki-light:#E36209;--shiki-default:#E36209;--shiki-dark:#FFAB70}html pre.shiki code .sIIMD, html code.shiki .sIIMD{--shiki-light:#032F62;--shiki-default:#032F62;--shiki-dark:#9ECBFF}html pre.shiki code .sBjJW, html code.shiki .sBjJW{--shiki-light:#005CC5;--shiki-default:#005CC5;--shiki-dark:#79B8FF}html .light .shiki span {color: var(--shiki-light);background: var(--shiki-light-bg);font-style: var(--shiki-light-font-style);font-weight: var(--shiki-light-font-weight);text-decoration: var(--shiki-light-text-decoration);}html.light .shiki span {color: var(--shiki-light);background: var(--shiki-light-bg);font-style: var(--shiki-light-font-style);font-weight: var(--shiki-light-font-weight);text-decoration: var(--shiki-light-text-decoration);}html .default .shiki span {color: var(--shiki-default);background: var(--shiki-default-bg);font-style: var(--shiki-default-font-style);font-weight: var(--shiki-default-font-weight);text-decoration: var(--shiki-default-text-decoration);}html .shiki span {color: var(--shiki-default);background: var(--shiki-default-bg);font-style: var(--shiki-default-font-style);font-weight: var(--shiki-default-font-weight);text-decoration: var(--shiki-default-text-decoration);}html .dark .shiki span {color: var(--shiki-dark);background: var(--shiki-dark-bg);font-style: var(--shiki-dark-font-style);font-weight: var(--shiki-dark-font-weight);text-decoration: var(--shiki-dark-text-decoration);}html.dark .shiki span {color: var(--shiki-dark);background: var(--shiki-dark-bg);font-style: var(--shiki-dark-font-style);font-weight: var(--shiki-dark-font-weight);text-decoration: var(--shiki-dark-text-decoration);}",{"title":11,"searchDepth":12,"depth":12,"links":1513},[1514,1515,1520,1521,1522],{"id":1212,"depth":12,"text":1213},{"id":1253,"depth":12,"text":1254,"children":1516},[1517,1518,1519],{"id":1258,"depth":305,"text":1258},{"id":1283,"depth":305,"text":1283},{"id":1279,"depth":305,"text":1279},{"id":1465,"depth":12,"text":1466},{"id":1497,"depth":12,"text":1498},{"id":158,"depth":12,"text":159},"cf208554-4904-4184-8ddb-05a8201c4347","My MCP server searches English and French website content, retrieves complete Markdown pages, and runs isolated JavaScript queries against page metadata.",[1526,1558,1566,1570],{"type":173,"id":1527,"name":796,"description":1528,"ecosystem":1529},"mcp-soubiran-dev","Expose the MCP server at `\u002Fmcp` and redirect other paths to the personal website.",[1530,1532,1547,1552],{"type":800,"id":1527,"name":809,"description":1531,"href":995},"Deploy the TypeScript Worker with Wrangler.",{"type":178,"id":1186,"name":179,"description":1533,"href":1200,"ecosystem":1534},"Host the Worker source, content tools, and deployment configuration.",[1535,1537,1539,1543],{"type":184,"name":1158,"description":1536,"href":1160},"Define the server, tool inputs, and protocol responses.",{"type":184,"name":1162,"description":1538,"href":1164},"Handle MCP requests at the Worker's `\u002Fmcp` endpoint.",{"type":184,"name":1540,"description":1541,"href":1542},"Zod","Validate localized page catalogs and tool inputs.","https:\u002F\u002Fzod.dev",{"type":184,"name":1544,"description":1545,"href":1546},"Acorn","Parse query code and require a single asynchronous arrow function.","https:\u002F\u002Fgithub.com\u002Facornjs\u002Facorn",{"type":212,"id":1548,"name":1549,"description":1550,"href":1551},"soubiran-dev-ai-search","Cloudflare AI Search","Search the `soubiran-dev` instance through the `AI_SEARCH` namespace binding.","https:\u002F\u002Fdevelopers.cloudflare.com\u002Fai-search\u002F",{"type":1553,"id":1554,"name":1555,"description":1556,"href":1557},"runtime","mcp-content-loader","Cloudflare Worker Loader","Execute metadata queries in temporary Workers with network access disabled.","https:\u002F\u002Fdevelopers.cloudflare.com\u002Fworkers\u002Fruntime-apis\u002Fworker-loader\u002F",{"type":212,"id":1559,"name":1560,"description":1561,"ecosystem":1562},"localized-website-catalogs","Localized page catalogs and Markdown","Load `pages.en.json` and `pages.fr.json`, then retrieve documents through each record's `markdownUrl`.",[1563],{"type":217,"name":76,"relationship":1564,"description":1565},"producer","Publish pages, posts, series, talks, and infrastructure documentation in a shared content catalog.",{"type":1567,"id":1568,"name":205,"description":1569,"href":207},"monitoring","mcp-sentry","Capture request and tool exceptions without collecting user information or HTTP bodies.",{"type":1567,"id":1571,"name":812,"description":1572,"href":813},"mcp-evlog","Record structured request and tool outcomes with sensitive content redacted.",{},"\u002Finfra\u002Fservices\u002Fmcp-soubiran-dev","---\ncontentId: cf208554-4904-4184-8ddb-05a8201c4347\ntitle: mcp.soubiran.dev\ndescription: >-\n  My MCP server searches English and French website content, retrieves complete\n  Markdown pages, and runs isolated JavaScript queries against page metadata.\nurl: 'https:\u002F\u002Fmcp.soubiran.dev\u002Fmcp'\nrepository: 'https:\u002F\u002Fgithub.com\u002Fbarbapapazes\u002Fmcp.soubiran.dev'\necosystem:\n  - type: deployment\n    id: mcp-soubiran-dev\n    name: Cloudflare Workers\n    description: Expose the MCP server at `\u002Fmcp` and redirect other paths to the personal website.\n    ecosystem:\n      - type: build\n        id: mcp-soubiran-dev\n        name: Wrangler\n        description: Deploy the TypeScript Worker with Wrangler.\n        href: 'https:\u002F\u002Fdevelopers.cloudflare.com\u002Fworkers\u002Fwrangler\u002F'\n      - type: repository\n        id: mcp.soubiran.dev\n        name: GitHub\n        description: Host the Worker source, content tools, and deployment configuration.\n        href: 'https:\u002F\u002Fgithub.com\u002Fbarbapapazes\u002Fmcp.soubiran.dev'\n        ecosystem:\n          - type: stack\n            name: Model Context Protocol SDK\n            description: Define the server, tool inputs, and protocol responses.\n            href: 'https:\u002F\u002Fmodelcontextprotocol.io'\n          - type: stack\n            name: Cloudflare Agents\n            description: Handle MCP requests at the Worker's `\u002Fmcp` endpoint.\n            href: 'https:\u002F\u002Fdevelopers.cloudflare.com\u002Fagents\u002F'\n          - type: stack\n            name: Zod\n            description: Validate localized page catalogs and tool inputs.\n            href: 'https:\u002F\u002Fzod.dev'\n          - type: stack\n            name: Acorn\n            description: Parse query code and require a single asynchronous arrow function.\n            href: 'https:\u002F\u002Fgithub.com\u002Facornjs\u002Facorn'\n      - type: data\n        id: soubiran-dev-ai-search\n        name: Cloudflare AI Search\n        description: Search the `soubiran-dev` instance through the `AI_SEARCH` namespace binding.\n        href: 'https:\u002F\u002Fdevelopers.cloudflare.com\u002Fai-search\u002F'\n      - type: runtime\n        id: mcp-content-loader\n        name: Cloudflare Worker Loader\n        description: Execute metadata queries in temporary Workers with network access disabled.\n        href: 'https:\u002F\u002Fdevelopers.cloudflare.com\u002Fworkers\u002Fruntime-apis\u002Fworker-loader\u002F'\n  - type: data\n    id: localized-website-catalogs\n    name: Localized page catalogs and Markdown\n    description: Load `pages.en.json` and `pages.fr.json`, then retrieve documents through each record's `markdownUrl`.\n    ecosystem:\n      - type: website\n        name: soubiran.dev\n        relationship: producer\n        description: Publish pages, posts, series, talks, and infrastructure documentation in a shared content catalog.\n  - type: monitoring\n    id: mcp-sentry\n    name: Sentry\n    description: Capture request and tool exceptions without collecting user information or HTTP bodies.\n    href: 'https:\u002F\u002Fsentry.io'\n  - type: monitoring\n    id: mcp-evlog\n    name: Evlog\n    description: Record structured request and tool outcomes with sensitive content redacted.\n    href: 'https:\u002F\u002Fevlog.dev'\n---\n\n[mcp.soubiran.dev](https:\u002F\u002Fmcp.soubiran.dev\u002Fmcp) gives assistants access to my public English and French content. It exposes three read-only tools for topic discovery, complete Markdown retrieval, and exact metadata queries.\n\nThe [TypeScript Worker](https:\u002F\u002Fgithub.com\u002Fbarbapapazes\u002Fmcp.soubiran.dev) uses the Model Context Protocol SDK and the Cloudflare Agents MCP handler. Requests to `\u002Fmcp` reach the server. Other paths redirect to [soubiran.dev](\u002Finfra\u002Fwebsites\u002Fsoubiran-dev). The handler does not require authentication.\n\n## One content catalog, two locales\n\nThe server loads the website's `\u002Fpages.en.json` and `\u002Fpages.fr.json` feeds in parallel. `PAGES_EN_URL` and `PAGES_FR_URL` configure their locations. Zod validates the records, and the loader combines them into one `pages` array.\n\nEach record has a stable ID, locale, content type, title, published URL, and `markdownUrl`. Descriptions, dates, and translation links are optional. Relative URLs are resolved against the feed URL. Duplicate IDs across the combined feeds cause an error rather than selecting one document arbitrarily.\n\nThe website catalog covers pages, posts, talks, series, series articles, and infrastructure documentation. [talks.soubiran.dev](\u002Finfra\u002Fwebsites\u002Ftalks-soubiran-dev) supplies talk content to the website during its build. The MCP reads the resulting website documents through the same catalog as the rest of the content.\n\nPublication stays with the website. The MCP fetches metadata and follows the supplied Markdown URLs; it does not maintain another document store.\n\n## Search, retrieve, or query\n\n### `search_content`\n\nThis tool accepts a natural-language query and searches the `soubiran-dev` Cloudflare AI Search instance. Search and catalog loading run in parallel. Retrieval requests up to 20 results with a match threshold of 0.6 and reranks them with `@cf\u002Fbaai\u002Fbge-reranker-base`.\n\nThe server matches each search chunk's metadata ID to a catalog record. It returns the score, excerpt, source key, and page metadata, including the published URL for citations. Chunks without a matching catalog entry are omitted.\n\nSearch finds relevant content, but ranking does not establish which page is newest. Exact date comparisons belong in `list_pages`, and an assistant should read the full document with `get_page` before summarizing it.\n\n### `get_page`\n\nThis tool accepts an exact stable ID, resolves it in the combined catalog, and fetches its `markdownUrl`. It returns the complete document as text. An unknown ID or failed upstream request produces an MCP tool error.\n\nA talk document can include its summary, event details, resource links, and transcript. The website composes that Markdown, so retrieval works the same way for a talk, post, or infrastructure page.\n\n### `list_pages`\n\nThis tool accepts a single asynchronous JavaScript arrow function. The function can read the combined `pages` array to filter by locale or type, match titles, compare dates, count records, or group metadata. It does not search document bodies.\n\nFor example, this query returns the five most recent dated English posts with the fields needed for retrieval and citation:\n\n```js\nasync () => pages\n  .filter(page => page.locale === 'en' && page.type === 'post' && page.date)\n  .sort((a, b) => b.date.localeCompare(a.date))\n  .slice(0, 5)\n  .map(({ id, title, date, url }) => ({ id, title, date, url }))\n```\n\nReturning only the relevant records and fields reduces how much metadata enters the model's context. Both discovery tools return IDs that `get_page` can resolve.\n\n## Isolated code execution\n\n`list_pages` uses Code Mode to answer metadata questions without a separate tool for every filter or aggregation. Acorn parses the submitted code and checks that it is one asynchronous arrow function. Input is limited to 20,000 characters.\n\nWorker Loader runs the function in a temporary Worker with a recursively frozen copy of `pages`. The Worker has no secrets, storage, or application bindings. Network access is disabled with `globalOutbound: null`, and the executor also disables `fetch`, `WebSocket`, and `EventSource`.\n\nExecution has a 50 ms CPU limit and permits no subrequests. The function must return a JSON-serializable value, and the server rejects results larger than 256 KiB. Query failures become tool errors rather than failing the MCP request handler.\n\n## Deployment and observability\n\nWrangler deploys the server to Cloudflare Workers with an AI Search namespace binding and a Worker Loader binding. The MCP handler uses a configurable allowed CORS origin.\n\nCloudflare persists logs and traces. Sentry captures exceptions, with user information and HTTP body collection disabled. Evlog records request duration, tool outcomes, selected upstream timings, result counts, and response sizes. Submitted query code is not logged. Redaction rules cover input and output payloads, queries, document content, excerpts, Markdown, authorization headers, and cookies.\n\n## Ecosystem\n\n::infra-ecosystem\n::\n",{"title":1186,"description":1524},{"loc":1574},"infra\u002Fservices\u002Fmcp-soubiran-dev","J1HgAzkigYvVb7k-sqa2CTdzeO4bcUiTSlYJMi4h7hI",{"id":1581,"title":151,"body":1582,"contentId":1718,"description":1719,"ecosystem":1720,"extension":17,"meta":1757,"navigation":19,"part":16,"path":150,"rawbody":1758,"repository":1730,"robots":16,"schemaOrg":16,"seo":1759,"sitemap":1760,"stem":1761,"url":1588,"__hash__":1762},"pages\u002Finfra\u002Fservices\u002Fredeploy-soubiran-dev.md",{"type":8,"value":1583,"toc":1712},[1584,1591,1602,1606,1613,1616,1619,1623,1634,1637,1648,1655,1675,1682,1686,1693,1696,1708,1710],[64,1585,1586,1590],{},[67,1587,151],{"href":1588,"rel":1589},"https:\u002F\u002Fredeploy.soubiran.dev",[71]," coordinates builds that depend on files published by another Cloudflare Worker. A GitHub Actions workflow sends the upstream Worker name and a downstream deploy hook. The service waits for the upstream deployment to reach production, then calls the hook.",[64,1592,1593,1594,1596,1597,1599,1600,246],{},"This keeps build-time dependencies in order across separate repositories. ",[67,1595,60],{"href":223}," must read current metadata from ",[67,1598,76],{"href":75},", while the personal website must read current talk metadata from ",[67,1601,1246],{"href":1245},[90,1603,1605],{"id":1604},"why-coordination-happens-outside-github-actions","Why coordination happens outside GitHub Actions",[64,1607,1608,1609,1612],{},"A push starts GitHub Actions and Cloudflare Builds independently. The source workflow knows that a commit reached ",[111,1610,1611],{},"main",", but it does not know when Cloudflare has promoted the resulting Worker version to production. Calling the dependent deploy hook immediately could start a build against the previous metadata.",[64,1614,1615],{},"I put that wait in one service instead of copying Cloudflare API checks into each repository. This fits projects that deploy independently but share a build-time data contract. Each source repository only declares what Worker must finish and which build should follow.",[64,1617,1618],{},"Cloudflare Workflows owns the long-running part. The request handler validates the input, creates a Workflow instance, and responds without holding the HTTP request open. The Workflow can retry deployment checks and retain its execution state independently of the caller.",[90,1620,1622],{"id":1621},"how-a-redeployment-flows","How a redeployment flows",[64,1624,868,1625,1633],{},[67,1626,1629,1632],{"href":1627,"rel":1628},"https:\u002F\u002Fgithub.com\u002Fbarbapapazes\u002Fredeploy.soubiran.dev\u002Fblob\u002Fmain\u002Fsrc\u002Findex.ts",[71],[111,1630,1631],{},"POST \u002Furl"," handler"," accepts a deploy hook URL and an optional Worker name. When the Worker name is absent, the Workflow calls the hook without waiting. Production callers include it because they need deployment ordering.",[64,1635,1636],{},"For a coordinated request, the Workflow first checks for another running instance with the same Worker and deploy hook. It hashes the hook URL when it builds the instance identifier, so it can reject duplicate work without putting the secret URL in that identifier.",[64,1638,868,1639,1647],{},[67,1640,1643,1646],{"href":1641,"rel":1642},"https:\u002F\u002Fgithub.com\u002Fbarbapapazes\u002Fredeploy.soubiran.dev\u002Fblob\u002Fmain\u002Fsrc\u002Fworkers.ts",[71],[111,1644,1645],{},"waitForLatestWorkersDeployment"," check"," compares two views from the Cloudflare API. It reads the latest build for the Worker and the build attached to its latest production deployment. The step succeeds only when the latest build completed successfully and both records point to the same build. Until then, Cloudflare Workflows retries the step.",[64,1649,1650,1651,1654],{},"Once those records match, the Workflow sends a ",[111,1652,1653],{},"POST"," request to the supplied deploy hook. The current flows are:",[1656,1657,1658,1668],"ul",{},[1659,1660,1661,1662,1664,1665,1667],"li",{},"A push to ",[111,1663,76],{}," waits for its Worker, then rebuilds ",[111,1666,60],{}," so website content, infrastructure documentation, and API data stay aligned.",[1659,1669,1661,1670,1664,1672,1674],{},[111,1671,1246],{},[111,1673,76],{}," with the new talk catalog.",[64,1676,1677,1678,1681],{},"The initial ",[111,1679,1680],{},"OK"," response confirms that the Workflow started. The Workflow execution records whether the later deployment check and deploy hook succeeded.",[90,1683,1685],{"id":1684},"access-and-operations","Access and operations",[64,1687,1688,1689,1692],{},"Cloudflare Access protects the endpoint with service tokens stored as GitHub Actions secrets. The Worker does not duplicate that authentication check. Callers also send an ",[111,1690,1691],{},"x-service"," header, which identifies the source repository in logs but does not grant access.",[64,1694,1695],{},"The deploy hook URL stays inside the Workflow payload. Structured events record its host and a redacted path, then connect the incoming request ID, caller name, Workflow instance, checked deployment, and hook result. This gives each cross-project deployment one trace without writing the hook secret to logs.",[64,1697,868,1698,1703,1704,1707],{},[67,1699,1702],{"href":1700,"rel":1701},"https:\u002F\u002Fgithub.com\u002Fbarbapapazes\u002Fredeploy.soubiran.dev\u002Fblob\u002Fmain\u002Fwrangler.jsonc",[71],"Wrangler configuration"," maps the custom domain to the Worker and binds the ",[111,1705,1706],{},"RedeploySoubiranDev"," Workflow class. Wrangler deploys both pieces together. GitHub Actions runs ESLint on pushes and pull requests. Cloudflare stores Workflow execution state and exposes the resulting events through its observability tools.",[90,1709,159],{"id":158},[161,1711],{},{"title":11,"searchDepth":12,"depth":12,"links":1713},[1714,1715,1716,1717],{"id":1604,"depth":12,"text":1605},{"id":1621,"depth":12,"text":1622},{"id":1684,"depth":12,"text":1685},{"id":158,"depth":12,"text":159},"867adb28-d1d1-4017-adca-ffe1972b5f19","The deployment coordinator that waits for an upstream Cloudflare Worker to reach production before it triggers a dependent build.",[1721,1751,1753,1755],{"type":791,"name":754,"description":1722,"ecosystem":1723},"Authenticate deployment requests before they reach the Worker.",[1724],{"type":173,"id":1725,"name":796,"description":1726,"ecosystem":1727},"redeploy-soubiran-dev","Validate each request and start its deployment workflow.",[1728,1740],{"type":178,"id":151,"name":179,"description":1729,"href":1730,"ecosystem":1731},"Host the public Worker source and deployment configuration.","https:\u002F\u002Fgithub.com\u002Fbarbapapazes\u002Fredeploy.soubiran.dev",[1732,1736,1738],{"type":184,"name":1733,"description":1734,"href":1735},"Cloudflare SDK","Read Worker build and production deployment state from the Cloudflare API.","https:\u002F\u002Fgithub.com\u002Fcloudflare\u002Fcloudflare-typescript",{"type":184,"name":812,"description":1737,"href":813},"Correlate the incoming request with its Workflow execution.",{"type":184,"name":809,"description":1739,"href":995},"Configure and deploy the Worker and Workflow binding.",{"type":815,"relationship":220,"name":245,"description":1741,"ecosystem":1742},"Wait for the upstream deployment and invoke the dependent deploy hook.",[1743],{"type":1168,"relationship":220,"name":1744,"description":1745,"ecosystem":1746},"Cloudflare deploy hooks","Start the dependent build after the upstream Worker reaches production.",[1747,1749],{"type":209,"relationship":220,"name":60,"description":1748},"Rebuild after infrastructure or personal website content changes.",{"type":217,"relationship":220,"name":76,"description":1750},"Rebuild after new talk metadata reaches production.",{"type":217,"name":76,"description":1752},"Request an API rebuild after publishing personal website content or infrastructure documentation.",{"type":217,"name":1246,"description":1754},"Request a personal website rebuild after publishing talk metadata.",{"type":818,"name":819,"description":1756},"Route `redeploy.soubiran.dev` to the Worker.",{},"---\ncontentId: 867adb28-d1d1-4017-adca-ffe1972b5f19\ntitle: redeploy.soubiran.dev\ndescription: >-\n  The deployment coordinator that waits for an upstream Cloudflare Worker to\n  reach production before it triggers a dependent build.\nurl: 'https:\u002F\u002Fredeploy.soubiran.dev'\nrepository: 'https:\u002F\u002Fgithub.com\u002Fbarbapapazes\u002Fredeploy.soubiran.dev'\necosystem:\n  - type: auth\n    name: Cloudflare One\n    description: Authenticate deployment requests before they reach the Worker.\n    ecosystem:\n      - type: deployment\n        id: redeploy-soubiran-dev\n        name: Cloudflare Workers\n        description: Validate each request and start its deployment workflow.\n        ecosystem:\n          - type: repository\n            id: redeploy.soubiran.dev\n            name: GitHub\n            description: Host the public Worker source and deployment configuration.\n            href: 'https:\u002F\u002Fgithub.com\u002Fbarbapapazes\u002Fredeploy.soubiran.dev'\n            ecosystem:\n              - type: stack\n                name: Cloudflare SDK\n                description: Read Worker build and production deployment state from the Cloudflare API.\n                href: 'https:\u002F\u002Fgithub.com\u002Fcloudflare\u002Fcloudflare-typescript'\n              - type: stack\n                name: Evlog\n                description: Correlate the incoming request with its Workflow execution.\n                href: 'https:\u002F\u002Fevlog.dev'\n              - type: stack\n                name: Wrangler\n                description: Configure and deploy the Worker and Workflow binding.\n                href: 'https:\u002F\u002Fdevelopers.cloudflare.com\u002Fworkers\u002Fwrangler\u002F'\n          - type: workflows\n            relationship: consumer\n            name: Cloudflare Workflows\n            description: Wait for the upstream deployment and invoke the dependent deploy hook.\n            ecosystem:\n              - type: ci\u002Fcd\n                relationship: consumer\n                name: Cloudflare deploy hooks\n                description: Start the dependent build after the upstream Worker reaches production.\n                ecosystem:\n                  - type: service\n                    relationship: consumer\n                    name: api.soubiran.dev\n                    description: Rebuild after infrastructure or personal website content changes.\n                  - type: website\n                    relationship: consumer\n                    name: soubiran.dev\n                    description: Rebuild after new talk metadata reaches production.\n  - type: website\n    name: soubiran.dev\n    description: Request an API rebuild after publishing personal website content or infrastructure documentation.\n  - type: website\n    name: talks.soubiran.dev\n    description: Request a personal website rebuild after publishing talk metadata.\n  - type: domain\n    name: Cloudflare Domains\n    description: Route `redeploy.soubiran.dev` to the Worker.\n---\n\n[redeploy.soubiran.dev](https:\u002F\u002Fredeploy.soubiran.dev) coordinates builds that depend on files published by another Cloudflare Worker. A GitHub Actions workflow sends the upstream Worker name and a downstream deploy hook. The service waits for the upstream deployment to reach production, then calls the hook.\n\nThis keeps build-time dependencies in order across separate repositories. [api.soubiran.dev](\u002Finfra\u002Fservices\u002Fapi-soubiran-dev) must read current metadata from [soubiran.dev](\u002Finfra\u002Fwebsites\u002Fsoubiran-dev), while the personal website must read current talk metadata from [talks.soubiran.dev](\u002Finfra\u002Fwebsites\u002Ftalks-soubiran-dev).\n\n## Why coordination happens outside GitHub Actions\n\nA push starts GitHub Actions and Cloudflare Builds independently. The source workflow knows that a commit reached `main`, but it does not know when Cloudflare has promoted the resulting Worker version to production. Calling the dependent deploy hook immediately could start a build against the previous metadata.\n\nI put that wait in one service instead of copying Cloudflare API checks into each repository. This fits projects that deploy independently but share a build-time data contract. Each source repository only declares what Worker must finish and which build should follow.\n\nCloudflare Workflows owns the long-running part. The request handler validates the input, creates a Workflow instance, and responds without holding the HTTP request open. The Workflow can retry deployment checks and retain its execution state independently of the caller.\n\n## How a redeployment flows\n\nThe [`POST \u002Furl` handler](https:\u002F\u002Fgithub.com\u002Fbarbapapazes\u002Fredeploy.soubiran.dev\u002Fblob\u002Fmain\u002Fsrc\u002Findex.ts) accepts a deploy hook URL and an optional Worker name. When the Worker name is absent, the Workflow calls the hook without waiting. Production callers include it because they need deployment ordering.\n\nFor a coordinated request, the Workflow first checks for another running instance with the same Worker and deploy hook. It hashes the hook URL when it builds the instance identifier, so it can reject duplicate work without putting the secret URL in that identifier.\n\nThe [`waitForLatestWorkersDeployment` check](https:\u002F\u002Fgithub.com\u002Fbarbapapazes\u002Fredeploy.soubiran.dev\u002Fblob\u002Fmain\u002Fsrc\u002Fworkers.ts) compares two views from the Cloudflare API. It reads the latest build for the Worker and the build attached to its latest production deployment. The step succeeds only when the latest build completed successfully and both records point to the same build. Until then, Cloudflare Workflows retries the step.\n\nOnce those records match, the Workflow sends a `POST` request to the supplied deploy hook. The current flows are:\n\n- A push to `soubiran.dev` waits for its Worker, then rebuilds `api.soubiran.dev` so website content, infrastructure documentation, and API data stay aligned.\n- A push to `talks.soubiran.dev` waits for its Worker, then rebuilds `soubiran.dev` with the new talk catalog.\n\nThe initial `OK` response confirms that the Workflow started. The Workflow execution records whether the later deployment check and deploy hook succeeded.\n\n## Access and operations\n\nCloudflare Access protects the endpoint with service tokens stored as GitHub Actions secrets. The Worker does not duplicate that authentication check. Callers also send an `x-service` header, which identifies the source repository in logs but does not grant access.\n\nThe deploy hook URL stays inside the Workflow payload. Structured events record its host and a redacted path, then connect the incoming request ID, caller name, Workflow instance, checked deployment, and hook result. This gives each cross-project deployment one trace without writing the hook secret to logs.\n\nThe [Wrangler configuration](https:\u002F\u002Fgithub.com\u002Fbarbapapazes\u002Fredeploy.soubiran.dev\u002Fblob\u002Fmain\u002Fwrangler.jsonc) maps the custom domain to the Worker and binds the `RedeploySoubiranDev` Workflow class. Wrangler deploys both pieces together. GitHub Actions runs ESLint on pushes and pull requests. Cloudflare stores Workflow execution state and exposes the resulting events through its observability tools.\n\n## Ecosystem\n\n::infra-ecosystem\n::\n",{"title":151,"description":1719},{"loc":150},"infra\u002Fservices\u002Fredeploy-soubiran-dev","IEM-ieE7kGruOuiR7UP0bljK4-xBRb_h3uCCT9jhiOI",{"id":1764,"title":76,"body":1765,"contentId":1906,"description":1907,"ecosystem":1908,"extension":17,"meta":1977,"navigation":19,"part":16,"path":75,"rawbody":1978,"repository":1776,"robots":16,"schemaOrg":16,"seo":1979,"sitemap":1980,"stem":1981,"url":1771,"__hash__":1982},"pages\u002Finfra\u002Fwebsites\u002Fsoubiran-dev.md",{"type":8,"value":1766,"toc":1899},[1767,1780,1784,1794,1801,1807,1810,1814,1824,1833,1842,1847,1851,1857,1863,1866,1872,1876,1889,1895,1897],[64,1768,1769,1773,1774,1779],{},[67,1770,76],{"href":1771,"rel":1772},"https:\u002F\u002Fsoubiran.dev",[71]," is my personal website. It brings posts, series, projects, talks, and infrastructure documentation into one Nuxt 4 application. The ",[67,1775,1778],{"href":1776,"rel":1777},"https:\u002F\u002Fgithub.com\u002Fbarbapapazes\u002Fbeta.soubiran.dev",[71],"source repository"," contains the content and the local Nuxt modules that implement the site's features.",[90,1781,1783],{"id":1782},"static-pages-build-time-content","Static pages, build-time content",[64,1785,1786,1789,1790,1793],{},[111,1787,1788],{},"nuxt generate"," produces the website in ",[111,1791,1792],{},".output\u002Fpublic",". Wrangler deploys that directory to Cloudflare Workers Static Assets, which serves the generated pages without running Nuxt for each visit.",[64,1795,1796,1797,1800],{},"Nuxt Content reads English and French Markdown into collections for pages, posts, and series. The infrastructure documentation lives in ",[111,1798,1799],{},"content\u002Finfra\u002F",", stays English-only, and uses the same page collection and shared layout. Its ecosystem and dataflow diagrams use Vue Flow.",[64,1802,1803,1804,1806],{},"The build also imports remote data. The projects section fetches GitHub repository metadata. The talks module reads both localized catalogs from ",[67,1805,1246],{"href":1245},", fetches each talk's metadata, and imports summaries and available transcripts into separate Nuxt Content collections. Talk pages combine those documents with recordings, slides, resource links, and event locations. The talk index includes a map and aggregate statistics.",[64,1808,1809],{},"Nuxt UI supplies interface components, with Vue and Tailwind CSS for the application code and styles. Local modules organize content, authentication, comments, reactions, feedback, analytics, and the assistant.",[90,1811,1813],{"id":1812},"a-catalog-for-readers-and-tools","A catalog for readers and tools",[64,1815,109,1816,114,1818,1820,1821,1823],{},[111,1817,113],{},[111,1819,117],{},". Each catalog is an array of records with a stable ID, locale, content type, title, description, published URL, and explicit ",[111,1822,1238],{},". Dated content includes a publication date. Translation links connect records available in another locale.",[64,1825,1826,1827,1829,1830,1832],{},"The catalogs include ordinary pages, infrastructure documentation, posts, series, series articles, and talks with summaries. Infrastructure documents use the ",[111,1828,1341],{}," type. Consumers follow ",[111,1831,1238],{}," to retrieve a document rather than infer its location from the HTML URL.",[64,1834,868,1835,1838,1839,246],{},[111,1836,1837],{},"\u002Fraw\u002F"," endpoints export readable Markdown. A talk document includes its event metadata, resource links, summary, and transcript when available. Generated documents for index pages include the corresponding content lists. The site also publishes RSS feeds and LLM-oriented content through ",[111,1840,1841],{},"nuxt-llms",[64,1843,1844,1846],{},[67,1845,1186],{"href":1574}," reads both page catalogs. The website therefore publishes the discovery metadata and Markdown used by the MCP tools, including the talks imported from the archive.",[90,1848,1850],{"id":1849},"interactions-belong-to-the-api","Interactions belong to the API",[64,1852,1853,1854,1856],{},"The generated pages do not own mutable user data. ",[67,1855,60],{"href":223}," handles accounts, comments, reactions, feedback, messages, and notifications. It also receives assistant requests and supplies the data used by the website's administration interface.",[64,1858,1859,1860,246],{},"Browser requests include session credentials. Writes use Laravel Sanctum's CSRF cookie, and Pinia Colada manages queries and mutations. Shared page components associate content interactions with a stable ",[111,1861,1862],{},"contentId",[64,1864,1865],{},"The assistant can include the current page as context and displays streamed responses, tool activity, and usage information. Its conversation requests go to the API rather than the static hosting Worker.",[64,1867,1868,1869,1871],{},"PartyKit handles the live viewer count separately. When a host is configured, the browser connects to the ",[111,1870,1266],{}," room and receives the active connection total.",[90,1873,1875],{"id":1874},"analytics-errors-and-metadata","Analytics, errors, and metadata",[64,1877,1878,1879,1882,1883,1888],{},"The shared ",[111,1880,1881],{},"useAnalytics"," composable sends page views and interaction events to ",[67,1884,1887],{"href":1885,"rel":1886},"https:\u002F\u002Fumami.soubiran.dev",[71],"Umami",". Events cover content links, language changes, talk resources, comments, reactions, feedback, assistant activity, and infrastructure diagram controls. Sentry captures errors and associates each release with the build's Git commit.",[64,1890,1891,1892,1894],{},"The site configuration defines ",[111,1893,1771],{}," as the canonical URL, English and French routes, a sitemap, a Person identity, and generated Open Graph images. Administration pages and authentication callbacks are excluded from indexing. Raw Markdown is excluded from the sitemap.",[90,1896,159],{"id":158},[161,1898],{},{"title":11,"searchDepth":12,"depth":12,"links":1900},[1901,1902,1903,1904,1905],{"id":1782,"depth":12,"text":1783},{"id":1812,"depth":12,"text":1813},{"id":1849,"depth":12,"text":1850},{"id":1874,"depth":12,"text":1875},{"id":158,"depth":12,"text":159},"6ea85ef4-62a7-49ab-b9c5-d1c28ae70808","My Nuxt personal website publishes posts, series, projects, localized talks, and infrastructure documentation, with static pages and API-backed interactions.",[1909,1942,1944,1948,1953,1955,1961,1967,1972,1975],{"type":173,"id":1910,"name":796,"description":1911,"ecosystem":1912},"beta-soubiran-dev","Serve the generated `.output\u002Fpublic` directory with Workers Static Assets.",[1913],{"type":800,"id":1910,"name":1914,"description":1915,"href":1916,"ecosystem":1917},"Nuxt generate","Generate static pages, localized catalogs, and Markdown endpoints, then deploy with Wrangler.","https:\u002F\u002Fnuxt.com\u002Fdocs\u002F4.x\u002Fapi\u002Fcommands\u002Fgenerate",[1918],{"type":178,"id":1919,"name":179,"description":1920,"href":1776,"ecosystem":1921},"beta.soubiran.dev","Host the personal website, Markdown content, and local Nuxt modules.",[1922,1925,1929,1930,1931,1934,1938],{"type":184,"name":1923,"href":1924},"Nuxt","https:\u002F\u002Fnuxt.com",{"type":184,"name":1926,"description":1927,"href":1928},"Nuxt Content","Import local Markdown and remote content into typed collections.","https:\u002F\u002Fcontent.nuxt.com",{"type":184,"name":1147,"href":1149},{"type":184,"name":1143,"href":1145},{"type":184,"name":1932,"href":1933},"Tailwind CSS","https:\u002F\u002Ftailwindcss.com",{"type":184,"name":1935,"description":1936,"href":1937},"Vue Flow","Render infrastructure ecosystem and dataflow diagrams.","https:\u002F\u002Fvueflow.dev",{"type":184,"name":1939,"description":1940,"href":1941},"Pinia Colada","Manage browser-side API queries and mutations.","https:\u002F\u002Fpinia-colada.esm.dev",{"type":217,"name":1246,"relationship":1564,"description":1943},"Supply localized talk catalogs, per-talk metadata, summaries, transcripts, and statistics at build time.",{"type":212,"id":1945,"name":1946,"description":1947},"infra-content","Infrastructure documentation","Read English-only Markdown from `content\u002Finfra\u002F` and publish it under `\u002Finfra`.",{"type":212,"id":1949,"name":1950,"description":1951,"href":1952},"github-projects","GitHub repository data","Fetch public repository metadata during the build for the projects section.","https:\u002F\u002Fdocs.github.com\u002Fen\u002Frest\u002Frepos\u002Frepos#get-a-repository",{"type":209,"name":60,"relationship":1564,"description":1954},"Handle authentication, comments, reactions, feedback, messages, notifications, administration, and assistant requests.",{"type":1956,"id":1957,"name":1958,"description":1959,"href":1960},"realtime","soubiran-dev-viewers","PartyKit","Connect to the `soubiran-dev` room for the live viewer count when a host is configured.","https:\u002F\u002Fwww.partykit.io",{"type":212,"relationship":220,"id":1962,"name":1560,"description":1963,"ecosystem":1964},"nuxt-pages-catalogs","Publish `pages.en.json`, `pages.fr.json`, and `\u002Fraw\u002F` Markdown documents with stable content IDs.",[1965],{"type":209,"relationship":220,"name":1186,"description":1966},"Load both catalogs to search, query metadata, and retrieve complete Markdown documents.",{"type":1968,"id":1969,"name":1887,"description":1970,"href":1971},"analytics","soubiran-dev-analytics","Track page views and user interactions through `useAnalytics`.","https:\u002F\u002Fumami.is",{"type":1567,"id":1973,"name":205,"description":1974,"href":207},"soubiran-dev-errors","Capture errors and identify releases by the build's Git commit.",{"type":818,"id":1266,"name":819,"description":1976},"Manage the website's domain.",{},"---\ncontentId: 6ea85ef4-62a7-49ab-b9c5-d1c28ae70808\ntitle: soubiran.dev\ndescription: >-\n  My Nuxt personal website publishes posts, series, projects, localized talks,\n  and infrastructure documentation, with static pages and API-backed interactions.\nurl: 'https:\u002F\u002Fsoubiran.dev'\nrepository: 'https:\u002F\u002Fgithub.com\u002Fbarbapapazes\u002Fbeta.soubiran.dev'\necosystem:\n  - type: deployment\n    id: beta-soubiran-dev\n    name: Cloudflare Workers\n    description: Serve the generated `.output\u002Fpublic` directory with Workers Static Assets.\n    ecosystem:\n      - type: build\n        id: beta-soubiran-dev\n        name: Nuxt generate\n        description: Generate static pages, localized catalogs, and Markdown endpoints, then deploy with Wrangler.\n        href: 'https:\u002F\u002Fnuxt.com\u002Fdocs\u002F4.x\u002Fapi\u002Fcommands\u002Fgenerate'\n        ecosystem:\n          - type: repository\n            id: beta.soubiran.dev\n            name: GitHub\n            description: Host the personal website, Markdown content, and local Nuxt modules.\n            href: 'https:\u002F\u002Fgithub.com\u002Fbarbapapazes\u002Fbeta.soubiran.dev'\n            ecosystem:\n              - type: stack\n                name: Nuxt\n                href: 'https:\u002F\u002Fnuxt.com'\n              - type: stack\n                name: Nuxt Content\n                description: Import local Markdown and remote content into typed collections.\n                href: 'https:\u002F\u002Fcontent.nuxt.com'\n              - type: stack\n                name: Nuxt UI\n                href: 'https:\u002F\u002Fui.nuxt.com'\n              - type: stack\n                name: Vue\n                href: 'https:\u002F\u002Fvuejs.org'\n              - type: stack\n                name: Tailwind CSS\n                href: 'https:\u002F\u002Ftailwindcss.com'\n              - type: stack\n                name: Vue Flow\n                description: Render infrastructure ecosystem and dataflow diagrams.\n                href: 'https:\u002F\u002Fvueflow.dev'\n              - type: stack\n                name: Pinia Colada\n                description: Manage browser-side API queries and mutations.\n                href: 'https:\u002F\u002Fpinia-colada.esm.dev'\n  - type: website\n    name: talks.soubiran.dev\n    relationship: producer\n    description: Supply localized talk catalogs, per-talk metadata, summaries, transcripts, and statistics at build time.\n  - type: data\n    id: infra-content\n    name: Infrastructure documentation\n    description: Read English-only Markdown from `content\u002Finfra\u002F` and publish it under `\u002Finfra`.\n  - type: data\n    id: github-projects\n    name: GitHub repository data\n    description: Fetch public repository metadata during the build for the projects section.\n    href: 'https:\u002F\u002Fdocs.github.com\u002Fen\u002Frest\u002Frepos\u002Frepos#get-a-repository'\n  - type: service\n    name: api.soubiran.dev\n    relationship: producer\n    description: Handle authentication, comments, reactions, feedback, messages, notifications, administration, and assistant requests.\n  - type: realtime\n    id: soubiran-dev-viewers\n    name: PartyKit\n    description: Connect to the `soubiran-dev` room for the live viewer count when a host is configured.\n    href: 'https:\u002F\u002Fwww.partykit.io'\n  - type: data\n    relationship: consumer\n    id: nuxt-pages-catalogs\n    name: Localized page catalogs and Markdown\n    description: Publish `pages.en.json`, `pages.fr.json`, and `\u002Fraw\u002F` Markdown documents with stable content IDs.\n    ecosystem:\n      - type: service\n        relationship: consumer\n        name: mcp.soubiran.dev\n        description: Load both catalogs to search, query metadata, and retrieve complete Markdown documents.\n  - type: analytics\n    id: soubiran-dev-analytics\n    name: Umami\n    description: Track page views and user interactions through `useAnalytics`.\n    href: 'https:\u002F\u002Fumami.is'\n  - type: monitoring\n    id: soubiran-dev-errors\n    name: Sentry\n    description: Capture errors and identify releases by the build's Git commit.\n    href: 'https:\u002F\u002Fsentry.io'\n  - type: domain\n    id: soubiran-dev\n    name: Cloudflare Domains\n    description: Manage the website's domain.\n---\n\n[soubiran.dev](https:\u002F\u002Fsoubiran.dev) is my personal website. It brings posts, series, projects, talks, and infrastructure documentation into one Nuxt 4 application. The [source repository](https:\u002F\u002Fgithub.com\u002Fbarbapapazes\u002Fbeta.soubiran.dev) contains the content and the local Nuxt modules that implement the site's features.\n\n## Static pages, build-time content\n\n`nuxt generate` produces the website in `.output\u002Fpublic`. Wrangler deploys that directory to Cloudflare Workers Static Assets, which serves the generated pages without running Nuxt for each visit.\n\nNuxt Content reads English and French Markdown into collections for pages, posts, and series. The infrastructure documentation lives in `content\u002Finfra\u002F`, stays English-only, and uses the same page collection and shared layout. Its ecosystem and dataflow diagrams use Vue Flow.\n\nThe build also imports remote data. The projects section fetches GitHub repository metadata. The talks module reads both localized catalogs from [talks.soubiran.dev](\u002Finfra\u002Fwebsites\u002Ftalks-soubiran-dev), fetches each talk's metadata, and imports summaries and available transcripts into separate Nuxt Content collections. Talk pages combine those documents with recordings, slides, resource links, and event locations. The talk index includes a map and aggregate statistics.\n\nNuxt UI supplies interface components, with Vue and Tailwind CSS for the application code and styles. Local modules organize content, authentication, comments, reactions, feedback, analytics, and the assistant.\n\n## A catalog for readers and tools\n\nThe website publishes `\u002Fpages.en.json` and `\u002Fpages.fr.json`. Each catalog is an array of records with a stable ID, locale, content type, title, description, published URL, and explicit `markdownUrl`. Dated content includes a publication date. Translation links connect records available in another locale.\n\nThe catalogs include ordinary pages, infrastructure documentation, posts, series, series articles, and talks with summaries. Infrastructure documents use the `page` type. Consumers follow `markdownUrl` to retrieve a document rather than infer its location from the HTML URL.\n\nThe `\u002Fraw\u002F` endpoints export readable Markdown. A talk document includes its event metadata, resource links, summary, and transcript when available. Generated documents for index pages include the corresponding content lists. The site also publishes RSS feeds and LLM-oriented content through `nuxt-llms`.\n\n[mcp.soubiran.dev](\u002Finfra\u002Fservices\u002Fmcp-soubiran-dev) reads both page catalogs. The website therefore publishes the discovery metadata and Markdown used by the MCP tools, including the talks imported from the archive.\n\n## Interactions belong to the API\n\nThe generated pages do not own mutable user data. [api.soubiran.dev](\u002Finfra\u002Fservices\u002Fapi-soubiran-dev) handles accounts, comments, reactions, feedback, messages, and notifications. It also receives assistant requests and supplies the data used by the website's administration interface.\n\nBrowser requests include session credentials. Writes use Laravel Sanctum's CSRF cookie, and Pinia Colada manages queries and mutations. Shared page components associate content interactions with a stable `contentId`.\n\nThe assistant can include the current page as context and displays streamed responses, tool activity, and usage information. Its conversation requests go to the API rather than the static hosting Worker.\n\nPartyKit handles the live viewer count separately. When a host is configured, the browser connects to the `soubiran-dev` room and receives the active connection total.\n\n## Analytics, errors, and metadata\n\nThe shared `useAnalytics` composable sends page views and interaction events to [Umami](https:\u002F\u002Fumami.soubiran.dev). Events cover content links, language changes, talk resources, comments, reactions, feedback, assistant activity, and infrastructure diagram controls. Sentry captures errors and associates each release with the build's Git commit.\n\nThe site configuration defines `https:\u002F\u002Fsoubiran.dev` as the canonical URL, English and French routes, a sitemap, a Person identity, and generated Open Graph images. Administration pages and authentication callbacks are excluded from indexing. Raw Markdown is excluded from the sitemap.\n\n## Ecosystem\n\n::infra-ecosystem\n::\n",{"title":76,"description":1907},{"loc":75},"infra\u002Fwebsites\u002Fsoubiran-dev","jtmuxJcLd8YhCZen0SM0uXByILT3JdtUvdHox1q5yLM",{"id":1984,"title":1246,"body":1985,"contentId":2161,"description":2162,"ecosystem":2163,"extension":17,"meta":2212,"navigation":19,"part":16,"path":1245,"rawbody":2213,"repository":2005,"robots":16,"schemaOrg":16,"seo":2214,"sitemap":2215,"stem":2216,"url":1991,"__hash__":2217},"pages\u002Finfra\u002Fwebsites\u002Ftalks-soubiran-dev.md",{"type":8,"value":1986,"toc":2154},[1987,1997,2001,2009,2016,2019,2023,2030,2056,2063,2080,2083,2087,2092,2098,2106,2111,2115,2134,2147,2150,2152],[64,1988,1989,1993,1994,1996],{},[67,1990,1246],{"href":1991,"rel":1992},"https:\u002F\u002Ftalks.soubiran.dev",[71]," publishes my conference and meetup slide decks. It also supplies the localized metadata, summaries, transcripts, and statistics used by ",[67,1995,76],{"href":75},". The archive's root redirects readers to the talk index on the personal website.",[90,1998,2000],{"id":1999},"one-workspace-per-appearance","One workspace per appearance",[64,2002,868,2003,2008],{},[67,2004,2007],{"href":2005,"rel":2006},"https:\u002F\u002Fgithub.com\u002Fbarbapapazes\u002Ftalks",[71],"public pnpm monorepo"," keeps a separate Slidev workspace for each event appearance. Each dated directory contains the slides and public assets for that presentation. Workspaces share dependency versions, root commands, and a local Slidev theme.",[64,2010,2011,2012,2015],{},"A talk's ",[111,2013,2014],{},"src\u002Fpackage.json"," defines its date, source language, topics, event, location coordinates, and resource links. English and French metadata each have a title, description, and globally unique UUID. The metadata locale is separate from the language spoken on stage, so an English page can describe a talk given in French.",[64,2017,2018],{},"The date and package name define the public deck path. The localized UUIDs identify the content in catalogs and in the personal website's interactions.",[90,2020,2022],{"id":2021},"validated-metadata-and-markdown","Validated metadata and Markdown",[64,2024,2025,2026,2029],{},"The build compiles the Slidev workspaces into a combined ",[111,2027,2028],{},"dist"," directory. Generation scripts validate the talk packages and publish:",[1656,2031,2032,2041,2050],{},[1659,2033,2034,114,2037,2040],{},[111,2035,2036],{},"talks.en.json",[111,2038,2039],{},"talks.fr.json",", arrays of localized talk records.",[1659,2042,2043,114,2046,2049],{},[111,2044,2045],{},"meta.en.json",[111,2047,2048],{},"meta.fr.json"," inside each talk's public directory.",[1659,2051,2052,2055],{},[111,2053,2054],{},"statistics.json",", with totals and counts by year, event, title, recording availability, and city.",[64,2057,2058,2059,2062],{},"Each localized record includes event details, topics, resource links, and artifact filenames. The shared ",[111,2060,2061],{},"@soubiran\u002Ftalks"," package defines the schemas and types used by the archive and the website.",[64,2064,2065,2066,114,2069,2072,2073,114,2076,2079],{},"Every talk must have ",[111,2067,2068],{},"summary.en.md",[111,2070,2071],{},"summary.fr.md",". A talk with a recording must also have ",[111,2074,2075],{},"transcript.en.md",[111,2077,2078],{},"transcript.fr.md",". Validation rejects missing required artifacts, transcripts attached to talks without recordings, duplicate locale IDs, and dates that do not match the event directory.",[64,2081,2082],{},"These checks run before the catalogs are written. Consumers can use the declared artifacts to decide which documents to fetch.",[90,2084,2086],{"id":2085},"publication-on-the-personal-website","Publication on the personal website",[64,2088,2089,2091],{},[67,2090,76],{"href":75}," fetches both localized catalogs during its build. It then reads per-talk metadata and imports summaries and transcripts into separate Nuxt Content collections. The generated talk pages display the summary, available transcript, recording, event location, and resource links.",[64,2093,2094,2095,2097],{},"The website reads ",[111,2096,2054],{}," for its statistics section and uses event coordinates for the talk map. Resource clicks and recording interactions are tracked through the website's analytics composable.",[64,2099,2100,2101,114,2103,2105],{},"Talks with summaries also appear in the website's ",[111,2102,113],{},[111,2104,117],{}," catalogs. Their Markdown documents combine event metadata, resources, the summary, and the available transcript.",[64,2107,2108,2110],{},[67,2109,1186],{"href":1574}," reads those website catalogs and retrieves the resulting Markdown. The MCP does not need to read the archive's talk catalog or select transcript files itself.",[90,2112,2114],{"id":2113},"assets-and-dependent-deployments","Assets and dependent deployments",[64,2116,2117,2118,346,2121,346,2124,346,2127,1488,2130,2133],{},"Thumbnails and localized Markdown artifacts live in each deck's public directory and ship with the Slidev build. PDF and audio exports use dedicated asset hosts. Generated redirects expose resource paths such as ",[111,2119,2120],{},"\u002Fpdf",[111,2122,2123],{},"\u002Faudio",[111,2125,2126],{},"\u002Frecording",[111,2128,2129],{},"\u002Ftranscript",[111,2131,2132],{},"\u002Fsrc",". The transcript redirect points readers to the talk page on the personal website.",[64,2135,2136,2137,2139,2140,2142,2143,2146],{},"Cloudflare Builds publishes the combined static output. After a push to ",[111,2138,1611],{},", GitHub Actions sends a request to ",[67,2141,151],{"href":150}," with the website's deploy hook and the ",[111,2144,2145],{},"talks"," Worker as the deployment to wait for. Cloudflare Access service credentials protect that request.",[64,2148,2149],{},"The website rebuild starts after the archive deployment completes, so it imports the published catalogs and artifacts rather than racing their deployment.",[90,2151,159],{"id":158},[161,2153],{},{"title":11,"searchDepth":12,"depth":12,"links":2155},[2156,2157,2158,2159,2160],{"id":1999,"depth":12,"text":2000},{"id":2021,"depth":12,"text":2022},{"id":2085,"depth":12,"text":2086},{"id":2113,"depth":12,"text":2114},{"id":158,"depth":12,"text":159},"a1a1a745-8ba1-47a9-9b02-9ed1f55fd3a8","The static archive for my Slidev decks publishes English and French metadata, summaries, transcripts, and statistics for my personal website.",[2164,2188,2198,2206,2210],{"type":173,"id":2145,"name":796,"description":2165,"ecosystem":2166},"Serve `dist`, including decks, thumbnails, localized catalogs, and Markdown artifacts.",[2167],{"type":800,"id":2145,"name":801,"description":2168,"ecosystem":2169},"Build the Slidev workspaces and deploy their combined static output.",[2170],{"type":178,"id":1246,"name":179,"description":2171,"href":2005,"ecosystem":2172},"Host the decks, shared theme, metadata package, and generation scripts.",[2173,2177,2180],{"type":184,"name":2174,"description":2175,"href":2176},"Slidev","Compile each event workspace into a static deck.","https:\u002F\u002Fsli.dev",{"type":184,"name":2061,"description":2178,"href":2179},"Define localized metadata schemas and statistics types shared with the website.","https:\u002F\u002Fwww.npmjs.com\u002Fpackage\u002F@soubiran\u002Ftalks",{"type":1168,"relationship":220,"name":1169,"description":2181,"ecosystem":2182},"Check changes and request a dependent website deployment after pushes to `main`.",[2183],{"type":209,"relationship":220,"name":151,"description":2184,"ecosystem":2185},"Wait for the `talks` Worker deployment, then invoke the website deploy hook.",[2186],{"type":217,"relationship":220,"name":76,"description":2187},"Rebuild using the published talk metadata and artifacts.",{"type":212,"relationship":220,"id":2189,"name":2190,"description":2191,"ecosystem":2192},"localized-talks-catalogs","Localized talk metadata and artifacts","Publish `talks.en.json`, `talks.fr.json`, per-talk metadata, summaries, and recorded-talk transcripts.",[2193],{"type":217,"relationship":220,"name":76,"description":2194,"ecosystem":2195},"Import both locales into Nuxt Content and publish talk pages and Markdown documents.",[2196],{"type":209,"relationship":220,"name":1186,"description":2197},"Discover and retrieve talk documents through the website's page catalogs.",{"type":212,"relationship":220,"id":2199,"name":2200,"description":2201,"href":2202,"ecosystem":2203},"talks-statistics","Talk statistics","Publish aggregate statistics in `statistics.json`.","https:\u002F\u002Ftalks.soubiran.dev\u002Fstatistics.json",[2204],{"type":217,"relationship":220,"name":76,"description":2205},"Render the statistics section from validated build-time data.",{"type":999,"id":2207,"name":2208,"description":2209},"talks-exports","PDF and audio storage","Serve exports through `pdf.talks.soubiran.dev` and `audio.talks.soubiran.dev`.",{"type":818,"name":819,"description":2211},"Route `talks.soubiran.dev` to the archive.",{},"---\ncontentId: a1a1a745-8ba1-47a9-9b02-9ed1f55fd3a8\ntitle: talks.soubiran.dev\ndescription: >-\n  The static archive for my Slidev decks publishes English and French metadata,\n  summaries, transcripts, and statistics for my personal website.\nurl: 'https:\u002F\u002Ftalks.soubiran.dev'\nrepository: 'https:\u002F\u002Fgithub.com\u002Fbarbapapazes\u002Ftalks'\necosystem:\n  - type: deployment\n    id: talks\n    name: Cloudflare Workers\n    description: Serve `dist`, including decks, thumbnails, localized catalogs, and Markdown artifacts.\n    ecosystem:\n      - type: build\n        id: talks\n        name: Cloudflare Build\n        description: Build the Slidev workspaces and deploy their combined static output.\n        ecosystem:\n          - type: repository\n            id: talks.soubiran.dev\n            name: GitHub\n            description: Host the decks, shared theme, metadata package, and generation scripts.\n            href: 'https:\u002F\u002Fgithub.com\u002Fbarbapapazes\u002Ftalks'\n            ecosystem:\n              - type: stack\n                name: Slidev\n                description: Compile each event workspace into a static deck.\n                href: 'https:\u002F\u002Fsli.dev'\n              - type: stack\n                name: '@soubiran\u002Ftalks'\n                description: Define localized metadata schemas and statistics types shared with the website.\n                href: 'https:\u002F\u002Fwww.npmjs.com\u002Fpackage\u002F@soubiran\u002Ftalks'\n              - type: ci\u002Fcd\n                relationship: consumer\n                name: GitHub Actions\n                description: Check changes and request a dependent website deployment after pushes to `main`.\n                ecosystem:\n                  - type: service\n                    relationship: consumer\n                    name: redeploy.soubiran.dev\n                    description: Wait for the `talks` Worker deployment, then invoke the website deploy hook.\n                    ecosystem:\n                      - type: website\n                        relationship: consumer\n                        name: soubiran.dev\n                        description: Rebuild using the published talk metadata and artifacts.\n  - type: data\n    relationship: consumer\n    id: localized-talks-catalogs\n    name: Localized talk metadata and artifacts\n    description: Publish `talks.en.json`, `talks.fr.json`, per-talk metadata, summaries, and recorded-talk transcripts.\n    ecosystem:\n      - type: website\n        relationship: consumer\n        name: soubiran.dev\n        description: Import both locales into Nuxt Content and publish talk pages and Markdown documents.\n        ecosystem:\n          - type: service\n            relationship: consumer\n            name: mcp.soubiran.dev\n            description: Discover and retrieve talk documents through the website's page catalogs.\n  - type: data\n    relationship: consumer\n    id: talks-statistics\n    name: Talk statistics\n    description: Publish aggregate statistics in `statistics.json`.\n    href: 'https:\u002F\u002Ftalks.soubiran.dev\u002Fstatistics.json'\n    ecosystem:\n      - type: website\n        relationship: consumer\n        name: soubiran.dev\n        description: Render the statistics section from validated build-time data.\n  - type: object-storage\n    id: talks-exports\n    name: PDF and audio storage\n    description: Serve exports through `pdf.talks.soubiran.dev` and `audio.talks.soubiran.dev`.\n  - type: domain\n    name: Cloudflare Domains\n    description: Route `talks.soubiran.dev` to the archive.\n---\n\n[talks.soubiran.dev](https:\u002F\u002Ftalks.soubiran.dev) publishes my conference and meetup slide decks. It also supplies the localized metadata, summaries, transcripts, and statistics used by [soubiran.dev](\u002Finfra\u002Fwebsites\u002Fsoubiran-dev). The archive's root redirects readers to the talk index on the personal website.\n\n## One workspace per appearance\n\nThe [public pnpm monorepo](https:\u002F\u002Fgithub.com\u002Fbarbapapazes\u002Ftalks) keeps a separate Slidev workspace for each event appearance. Each dated directory contains the slides and public assets for that presentation. Workspaces share dependency versions, root commands, and a local Slidev theme.\n\nA talk's `src\u002Fpackage.json` defines its date, source language, topics, event, location coordinates, and resource links. English and French metadata each have a title, description, and globally unique UUID. The metadata locale is separate from the language spoken on stage, so an English page can describe a talk given in French.\n\nThe date and package name define the public deck path. The localized UUIDs identify the content in catalogs and in the personal website's interactions.\n\n## Validated metadata and Markdown\n\nThe build compiles the Slidev workspaces into a combined `dist` directory. Generation scripts validate the talk packages and publish:\n\n- `talks.en.json` and `talks.fr.json`, arrays of localized talk records.\n- `meta.en.json` and `meta.fr.json` inside each talk's public directory.\n- `statistics.json`, with totals and counts by year, event, title, recording availability, and city.\n\nEach localized record includes event details, topics, resource links, and artifact filenames. The shared `@soubiran\u002Ftalks` package defines the schemas and types used by the archive and the website.\n\nEvery talk must have `summary.en.md` and `summary.fr.md`. A talk with a recording must also have `transcript.en.md` and `transcript.fr.md`. Validation rejects missing required artifacts, transcripts attached to talks without recordings, duplicate locale IDs, and dates that do not match the event directory.\n\nThese checks run before the catalogs are written. Consumers can use the declared artifacts to decide which documents to fetch.\n\n## Publication on the personal website\n\n[soubiran.dev](\u002Finfra\u002Fwebsites\u002Fsoubiran-dev) fetches both localized catalogs during its build. It then reads per-talk metadata and imports summaries and transcripts into separate Nuxt Content collections. The generated talk pages display the summary, available transcript, recording, event location, and resource links.\n\nThe website reads `statistics.json` for its statistics section and uses event coordinates for the talk map. Resource clicks and recording interactions are tracked through the website's analytics composable.\n\nTalks with summaries also appear in the website's `\u002Fpages.en.json` and `\u002Fpages.fr.json` catalogs. Their Markdown documents combine event metadata, resources, the summary, and the available transcript.\n\n[mcp.soubiran.dev](\u002Finfra\u002Fservices\u002Fmcp-soubiran-dev) reads those website catalogs and retrieves the resulting Markdown. The MCP does not need to read the archive's talk catalog or select transcript files itself.\n\n## Assets and dependent deployments\n\nThumbnails and localized Markdown artifacts live in each deck's public directory and ship with the Slidev build. PDF and audio exports use dedicated asset hosts. Generated redirects expose resource paths such as `\u002Fpdf`, `\u002Faudio`, `\u002Frecording`, `\u002Ftranscript`, and `\u002Fsrc`. The transcript redirect points readers to the talk page on the personal website.\n\nCloudflare Builds publishes the combined static output. After a push to `main`, GitHub Actions sends a request to [redeploy.soubiran.dev](\u002Finfra\u002Fservices\u002Fredeploy-soubiran-dev) with the website's deploy hook and the `talks` Worker as the deployment to wait for. Cloudflare Access service credentials protect that request.\n\nThe website rebuild starts after the archive deployment completes, so it imports the published catalogs and artifacts rather than racing their deployment.\n\n## Ecosystem\n\n::infra-ecosystem\n::\n",{"title":1246,"description":2162},{"loc":1245},"infra\u002Fwebsites\u002Ftalks-soubiran-dev","dJldPFiO04nt5iHtBlcP5RMtU5u-F8GGPNuLIatFJjg",1795016656290]