YouTube readiness · pkg_01KSHR025HW9EXD2VN094Y6TDP

Can I post this to YouTube? Yes — and it can upload itself.

Every YouTube field for the Quiet CPU Coolers package is drafted and sitting at ready_for_review. Content readiness is complete. The only question left is the route: ChannelHelm can now push the video for you via the Direct YouTube Data API (once the brand is connected), hand Shorts to Zernio, or you can paste fields into YouTube Studio by hand. This page maps each asset to its status and shows the one switch that ships it.

The verdict

✓ All YouTube content is generated and approval-ready.
✓ Direct upload is built — connect the brand's YouTube OAuth and set youtube_dispatch_target='youtube_direct' to auto-publish.
→ Not yet connected on this brand: until then, approving routes to manual (you paste into Studio).

Readiness at a glance

Content: ready. Route: one switch away.

The gauge scores two independent dimensions. Content readiness (are all the required assets present?) is at 100%. Dispatch readiness (is an automated route configured?) waits on the brand's OAuth connection.

CONTENT READINESS 100% 6 / 6 required assets present DISPATCH READINESS (this brand) manual no OAuth → falls back to paste Bridge: Connect YouTube on /brands/[id] + set youtube_dispatch_target = youtube_direct → auto-upload

Content is fully ready; only the route is unconfigured on this brand. The Direct path requires GOOGLE_OAUTH_CLIENT_ID/SECRET in /settings and a per-brand connect.

Asset completeness

Every required field, mapped to its status.

The Direct upload bundles youtube_title_set (the trigger) with youtube_description, youtube_chapters, youtube_tags, and the top thumbnail_concept (an AI-generated image, with frame-extract fallback). All sit at ready_for_review — one approve away from approved, then dispatched, then published.

draft ready_for_review approved dispatched published youtube_title_set trigger youtube_description v2 prompt youtube_chapters youtube_tags 15 scored thumbnail_concept 2 AI concepts transcript clean short_clip_plan → Zernio after render plan only — needs clip_render → rendered_short_clip ready_for_review 6 long-form assets — content complete short_clip_plan not rendered yet — Shorts route via Zernio remaining path approve → dispatch → publish (one click on Direct)

Solid amber = where each asset is now (ready_for_review). Dashed = the enum it still has to traverse. On the Direct path, approving youtube_title_set carries its siblings to dispatched automatically.

Package snapshot

What's in the box.

6 long-form YouTube assets + thumbnails + clean transcript, plus social drafts — all at ready_for_review from Codex CLI (gpt-5.5).

Source

Best Quiet CPU Coolers for Sustained AI Workloads
~8 min · uploaded_video

Brand

Thorsten Meyer AI
YT channel UCEqc9louhpbD4Ff0eOozFRQ ✓

Profile

standard_audio_visual
all 4 layers ran

YouTube fields

4 ready_for_review

title_set · description · chapters · tags

Thumbnails

2 AI concepts

AI-generated images · plain + headline variants — both ranked (frame-extract fallback if no image provider)

YouTube Shorts

plan only

short_clip_plan not yet rendered · ships via Zernio

Other platforms

5 drafted

linkedin · x_post · x_thread · article_brief · newsletter

YouTube Direct API

built

worker uploads via Data API v3 once connected

This brand's connection

not connected

youtube_oauth empty · target = manual

Three ways to ship

One needs no setup. One uploads for you. One handles Shorts.

The brand's youtube_dispatch_target decides. Full comparison on the publish options page.

Direct API recommended

ChannelHelm uploads original.mp4 straight to YouTube. Privacy defaults to private; the per-package picker also offers public / unlisted / schedule.

  • Set GOOGLE_OAUTH_CLIENT_ID/SECRET in /settings (one client, all brands).
  • /brands/[id]Connect YouTube → consent → token saved encrypted.
  • Set target to youtube_direct, then approve — the worker does the rest.

Zernio for Shorts

The only v1 path for Shorts. After clip_render, a rendered_short_clip goes to TikTok + Instagram + YouTube Shorts in one createPost.

  • Set ZERNIO_API_KEY, connect networks, paste each acc_… on the brand.
  • Needs a signed-media surface (CLOUDFLARE_TUNNEL_HOSTNAME + MEDIA_URL_SECRET).
  • privacy='schedule' + publish_at → Zernio scheduledFor.

Manual default

Always available, zero setup. Copy each field, upload the MP4 yourself. The dispatch is recorded for audit but makes no API call.

  • Approve the YouTube assets — each card has a Copy button.
  • Upload at studio.youtube.com; paste title / description / tags.
  • Pick an AI thumbnail (plain or headline variant); settles at dispatched (no published webhook).
The five title candidates

Top pick is the obvious winner.

Scores are model-judged hook strength (0–100). The Direct upload uses the title at selectedIndex from the youtube_title_set payload — open the Studio to swap.

93
Quiet CPU Coolers for AI Workstations in 2026
91
Why AI Workstations Need Marathon Cooling
88
Undervolt First: Quiet Cooling for AI Rigs
86
Stop Chasing Peak Benchmarks for AI Cooling
84
Air vs AIO: Cooling Trade-Offs for AI Rigs
Paste-ready preview

What the upload carries.

On the Direct path these fields go to videos.insert automatically. On the manual path, this is exactly what you paste into YouTube Studio.

YT Upload · Details (preview rendering)
Quiet CPU Coolers for AI Workstations in 2026
Your AI workstation should not sound like a server room trying to take off. The real question is not just how much performance you can buy, but how much quiet sustained performance you can live with every day. In this explainer, we break down why AI workloads punish coolers differently than gaming, why undervolting and power caps come before new hardware, and how sustained heat changes the air vs AIO liquid cooling decision. We also walk through 2026 quiet CPU cooler picks, from value air coolers to high-headroom AIOs and fanless specialist builds, then tie it all together with a practical framework: heat output, noise tolerance, fit, airflow, and long-term reliability. 🛠️ Chapters 0:00 Build a quiet high-performance AI rig 0:40 Gaming spikes vs AI marathon loads 1:44 Undervolt before buying a cooler 2:16 Sustained heat, big fans, low noise 2:56 Air reliability vs AIO headroom 3:28 2026 quiet cooler picks overview 4:24 Value and near-silent air coolers 5:20 AIO and fanless options for edge cases 6:24 Fit, airflow, and the final build order Like the video and subscribe to {{brand.name}} for more practical AI workstation and hardware breakdowns. Would you trade peak benchmark numbers for an AI rig you can barely hear? 🔇 #AI #AIWorkstation #CPUCooling #PCBuilding #QuietPC #AIOCooling #Undervolting #Hardware
0:00 — Quiet cooling goals for AI rigs 0:40 — Gaming spikes versus AI marathons 1:44 — Undervolt and power cap first 2:16 — Sustained heat and air versus AIO 3:28 — Top quiet CPU coolers for 2026 3:52 — Noctua NH-D55 premium air cooling 4:24 — Peerless Assassin and Dark Rock picks 5:20 — AIO and fanless options for edge cases 6:24 — Choose by heat, noise, fit, airflow 7:20 — Three-step cooling order of operations
quiet CPU cooler AI workstation cooling CPU cooling for AI silent AI PC build best CPU cooler 2026 sustained CPU load undervolting CPU air vs AIO cooling AIO liquid cooling Thermalright Peerless Assassin Noctua NH-D55 Arctic Liquid Freezer 3 Pro 360 Be Quiet Dark Rock Pro 5 Ryzen 9 cooling Core i9 cooling
Default: private — nothing goes public by accident.
Per-package picker also offers public · unlisted · schedule (publish_at, auto-flips public).
AI-generated images: the LLM drafts distinct visual concepts + ≤4-word headlines, the image provider (Runware / Flux-Z-Image) renders each, and ffmpeg cuts a plain + headline-overlay variant per concept. You pick one; it auto-uploads via thumbnails.set on the Direct path. With no image provider configured, this falls back to frame extraction at the top hook timestamp.
media/Thorsten Meyer AI/src_01KSHR023MRZZ3AHZQ49YM29Y7/thumbs/concept_01.jpg
media/Thorsten Meyer AI/src_01KSHR023MRZZ3AHZQ49YM29Y7/original.mp4
The Direct path, end to end

Ship it in five steps.

Connect the brand (one-time)

Paste GOOGLE_OAUTH_CLIENT_ID/SECRET in /settings, then /brands/[id]Connect YouTube → consent. The refresh token saves encrypted on brands.youtube_oauth. Set youtube_dispatch_target='youtube_direct'.

Open the package in the Studio

/packages/pkg_01KSHR025HW9EXD2VN094Y6TDP — pick the top title (or override the selectedIndex), and choose visibility in the per-package picker.

Approve youtube_title_set

That single approve enqueues a dispatch job. The worker bundles youtube_description / youtube_chapters / youtube_tags + the top thumbnail, then uploads — flipping all the siblings to dispatched in the same transaction.

The worker uploads

Refresh → access token → videos.insert (resumable, streamed from disk) → thumbnails.set. The YouTube URL is mirrored onto packages.intelligence.published.youtube and the red ▶ youtu.be/… chip lights up.

Shorts (separately, via Zernio)

Approve the short_clip_planclip_render produces rendered_short_clip → approving that dispatches to TikTok + Instagram + YouTube Shorts in one Zernio call. Direct Shorts is deferred to v2.

Two things worth fixing before you upload

Small bugs I spotted.

⚠ Package status drift

DB says packages.status = 'analyzed' but every asset is ready_for_review. The markReadyForReviewIfComplete lifecycle bumper didn't fire — probably because the worker got killed during a pnpm dev:all restart cycle while the last generate_asset was completing.

Effect: the dashboard filters may not list this package under “ready” yet. Fix: one-line SQL or a 5-line lifecycle-resync. Say the word.

⚠ Unresolved template token in description

The description contains the literal string {{brand.name}} instead of “Thorsten Meyer AI”. The prompt's mustache placeholder wasn't substituted at generate-time.

Effect: the upload would carry “subscribe to {{brand.name}}”. Fix: patch workers/lib/generate.ts (brand interpolation) + regenerate that one asset before dispatching.

Open it.

The Studio is where you approve and pick visibility. Connect the brand once and the next approve uploads itself — or compare all three routes first.

▶ Open package in Studio compare publish options