Skill · v1.0

Social Batch Queue

Social Batch Queue is a skill for shipping an approved 9-post batch as scheduled posts. Image generation, asset staging, scheduler integration with channel discovery and slot resolution, post-publish handoff. The full pipeline lives here. Pairs with Social Batch Drafting.

Install

Claude Code (CLI / WSL / Git Bash)

/plugin marketplace add https://www.infinitegameos.io/marketplace.json
/plugin install social-batch-queue@igos-library

Claude Code (VS Code)

Install in VS Code

Opens the Claude Code plugins dialog with the marketplace and skill prefilled. Requires the Claude Code VS Code extension installed and signed in. Or paste the snippet below into .claude/settings.json for VS Code, JetBrains or any setup that prefers manual config.

{
  "extraKnownMarketplaces": {
    "igos-library": {
      "source": {
        "source": "url",
        "url": "https://www.infinitegameos.io/marketplace.json"
      }
    }
  },
  "enabledPlugins": {
    "social-batch-queue@igos-library": true
  }
}

Direct markdown URL (Claude Code, Cursor, Codex CLI)

https://www.infinitegameos.io/markdown/skills/social-batch-queue

Cursor (.mdc rules file)

curl -O https://www.infinitegameos.io/install/cursor/social-batch-queue.mdc

Aider, Cline, any agent with --read

curl -O https://www.infinitegameos.io/markdown/skills/social-batch-queue
aider --read social-batch-queue.md

Definition

Social Batch Queue takes an approved batch markdown file (status `approved-pending-scheduling`) and ships it as scheduled posts across the operator's social platforms. Image generation runs first: a cinematic Day 3 background per the brief in the batch file, plus headline replacements on the Day 1 hook template and the Day 3 action template, plus bulk text replacement across the 8 carousel slides for Day 2. A Desktop review pause surfaces all rendered exports for operator review before any asset staging or scheduler queue runs. After approval, assets stage to a public URL host, channel IDs discover at run-time (never hardcoded), schedules resolve against each channel's existing slots and posts queue with alt text on every attachment. The output is a batch file with execution log entries, scheduled posts in the operator's scheduler with permalinks captured and a structured manual-handoff list for any platform-native operations the scheduler can't do.

Reference implementations and substitutability

The skill names a specific stack as the reference implementation, then frames each component generically. The reference image-generation pipeline uses fal.ai for backgrounds, Canva for template editing, Vercel Blob for public asset staging and a cloud-synced drive for local archive. The reference scheduler is Buffer with channel discovery and custom-scheduled posting. Operators with different stacks substitute component by component: any image-generation pipeline that produces PNG and PDF outputs and stages them at a public URL works for the asset side; any scheduling tool with channel discovery, slot configuration and a `customScheduled` API works for the scheduler side.

The structural discipline is what holds across stacks. Desktop review pause as the operator gate. Font sizes locked at template-load time. Vertical centering check after every text replacement on single-headline templates. Heading character budgets enforced at draft time so font sizes never have to drift. Channel IDs discovered at run-time. Native-isolation test as the cheapest first diagnostic on scheduler failure.

The Desktop review pause

Image generation is where review prevents downstream rework. A centering issue caught in the Desktop folder is cheap to fix; the same issue caught after the queue runs is expensive (re-export, re-upload, re-queue, possibly delete and re-create posts). The pause stays default-on for the first several live runs of this skill on any given operator stack. After multiple consecutive perfect first-time renders, the pause may relax to default-off as trust builds.

When the pause is active, the skill surfaces the Desktop folder path and the file inventory, then waits for explicit "approved" or specific adjustment requests. On adjustment, the skill loops back to the rendering step for the affected images. This is the firm gate. Bypassing the pause when it is set to true defeats the trust-building rhythm and reintroduces the rework cost the pause exists to prevent.

Use Cases

First live run on a new operator stack

An operator runs the skill for the first time. Desktop review pause is default-on. The skill renders all 10 image exports (Day 1 hook, 8 carousel slides plus Day 3 action), surfaces the Desktop folder for review and waits. The operator catches a font-size drift on slide 6 and a centering shift on Day 3. The skill loops back to the affected renders, fixes both and re-surfaces. Approved. Assets stage, schedule and queue.

Multi-batch arc execution with scheduler tier awareness

An operator runs three batches across a campaign window. The scheduler is on a free tier with a 9-slot queue cap. The skill spaces the batches so each batch's queue is cleared (or nearly cleared) before the next queues. Channel ID discovery runs every batch since the scheduler rotates IDs on disconnect.

Image-pipeline-only run for asset reuse

An operator wants the rendered assets but plans to schedule manually inside the platform-native UI. The skill runs Steps 1 through 5 (read draft, generate background, render templates, Desktop review, stage to public URLs) and then halts. The operator pulls the URLs and posts manually. The voice-and-copy boundary holds; the operator never edits the approved batch file.

Scheduler-failure recovery with native-isolation test

A scheduled post fails to publish on Instagram. The skill halts before iterating scheduler-layer variations. It surfaces the native-isolation test directive: post directly on the platform first as the cheapest diagnostic. If the native post succeeds, the issue is scheduler-layer; if it fails, the issue is platform-side. The test costs minutes and prevents hours of fruitless scheduler retries.

Substituting the reference image stack with the operator's preferred tools

An operator already has a Figma-based template editing flow and a Stability AI image generation pipeline. The skill's structural discipline (Desktop review, font-size lock, vertical centering check, character budgets, parallel-download race-condition discipline) applies to the operator's tools without modification. The reference stack is a starting point, not a requirement.

FAQ

Does this skill require Buffer specifically?

No. Buffer is the reference implementation; the skill names it explicitly so operators can see the working pattern. Any scheduling tool with channel discovery, slot configuration and a `customScheduled` API works. The structural discipline (channel IDs discovered at run-time, native-isolation test on failure, alt text on every attachment) holds across schedulers.

What happens if the batch draft contains voice violations?

The skill halts and routes back to the Drafting skill. Voice changes are out of scope here; this skill executes copy that Drafting has already approved. Publishing a draft with voice violations would defeat the Drafting skill's gate. The boundary is firm.

Why is the Desktop review pause non-negotiable for first runs?

Image-generation issues caught in the Desktop folder are cheap (re-export and continue). The same issues caught after the queue runs are expensive (re-export, re-upload, re-queue, possibly delete and re-create). The pause is the operator's leverage point. After multiple consecutive perfect renders, the pause may relax; until then it stays default-on.

Can I use a different image-generation stack?

Yes. The skill names fal.ai, Canva and Vercel Blob as the reference implementation. The image-generation pipeline is component-by-component substitutable. The structural discipline is what travels across stacks: Desktop review, font-size lock, vertical centering check, heading character budgets, parallel-download race-condition discipline.

How does the skill handle scheduler-to-platform failures?

On any scheduler error, the skill halts iteration and surfaces the native-isolation test as the cheapest first diagnostic: post directly on the platform first to determine whether the issue is scheduler-layer or platform-side. The skill does not iterate scheduler-layer variations beyond two or three attempts without operator authorization.

Related

Social Batch Queue pairs with Social Batch Drafting. Drafting produces the approved batch; Queue ships it as scheduled posts. The two work as a flywheel. The Sovereign Life Playbook is the upstream design frame for which work is worth distributing in the first place.

See the Sovereign Life Playbook