Concept PDF
The full architecture of an AI-driven shorts pipeline. Read in 90 minutes; build in a weekend.
$50
// source-owned · multi-channel · module-ready
Read the system map first: AI mode, manual mode, clip farming, TikTok publishing, captions, analytics, and rulesets.
// pick your entry point
The full architecture of an AI-driven shorts pipeline. Read in 90 minutes; build in a weekend.
$50
PDF + full repo with AI mode enabled. LLM-driven script + metadata generation. Bring your own AI API key.
$250
PDF + AI repo + all shipped module add-ons, including the n8n Workflow Pack and AI Rulesets. Saves $251 vs à la carte.
$1,000
$749
Save $251
// mass-production SaaS stack
A rented setup that gets closer to daily production needs higher-volume subscriptions for generation, clipping, voice, captions, and repurposing. agentshorts keeps the bundle as source code you own; provider usage and API costs remain yours.
one-year production SaaS stack costs about $4,063 more than the all-in bundle. bundle saves $251.
This delta is purchase-price math, not an all-in operating-cost promise. The repo can still use adjacent usage-billed model and voice providers through your own keys, so generation usage remains a pass-through cost you control.
Evidence uses public paid-plan prices checked in May 2026 for production-leaning creator/team tiers. Taxes, overages, FX, API usage, and plan changes are not included.
// two modes · same pipeline
AI mode uses your Anthropic key for scripts, metadata, and localization while the same repo handles TTS, video assembly, captions, upload prep, and review gates.
You write scripts and metadata in markdown. The downstream system still handles voice, video, captions, translation, YouTube upload, and TikTok handoff.
Generated videos are configurable from source: presets, aspect ratios, captions, colors, templates, voices, policy checks, and export rules live in files you can inspect and change.
// tier-3 modules
Channel performance dashboards, bulk underperformer cleanup, schedule intelligence.
$100 standalone
or $749 all-in (save $251)
Twitch + YouTube viral-clip discovery, 9:16 conversion, multi-platform syndication.
$100 standalone
or $749 all-in (save $251)
Direct TikTok Content Posting API integration, multi-account fanout, retry logic.
$100 standalone
or $749 all-in (save $251)
Retention-tuned caption rendering, viral-mode style presets, live preview.
$100 standalone
or $749 all-in (save $251)
Importable n8n AI workflow automation templates for intake, approvals, render-worker handoff, publishing queues, and analytics callbacks.
$100 standalone
or $749 all-in (save $251)
YouTube July-2025 inauthentic-content checklist, Made-with-AI label helper, title/desc/tag char limits, policy-risk heuristics, advertiser-friendly scoring.
$100 standalone
or $749 all-in (save $251)
TikTok AI label helper (C2PA-aware), FYP-risk heuristics, caption + hashtag rules, music license cross-post check.
$100 standalone
or $749 all-in (save $251)
// bundle calculator
full solution à la carte: $1,000 · bundle: $749 · you save: $251
Architecture
Seven file-backed stages. AI or manual inputs. Modules hook in at the boundaries.
# 1. Pick the operating mode (config/profile.yml)mode: "ai" # or "manual" — same downstream pipelinegeneration: preset: "starter"languages: targets: ["en", "pt-br"] # 2. Verify config, auth TikTok, and discover installed modules$ node doctor.mjs$ node tiktok-auth.mjs horror-en$ node bin/agentshorts.mjs --list # 3. Produce clips and publish through the same upload boundary$ node produce-clips-week.mjs --shorts-per-channel 1$ node upload-clips-week.mjs [pipeline] scan → script/manual → tts → video → metadata → upload [modules] viral-captions:before:upload → ai-rulesets-youtube:before:upload [tiktok] inbox draft created for operator reviewWhat you buy is a Node repo you clone, configure, and run. The lines below mirror the current buyer repo: AI/manual modes, TikTok auth, clip batches, and module discovery. See pricing →
Assumption-based monthly revenue for a faceless channel running the pipeline daily. Move the sliders. The chart updates.
Monetization tier
Higher tier = better niche, better hook, multi-platform monetization, affiliate income, sponsorships.
Modeled monthly revenue
$375
Year 1 cumulative
$3,694
Full AI pipeline price recovered in scenario
3 months
$250 of running
Cumulative income — 12 months
USD// rotating workflow notes
Illustrative scenario notes, not buyer reviews or ratings. English storefront notes stay English-only. Real outcomes depend on niche, inputs, review discipline, and distribution.
US · podcast producer
generated scenario
It did not pop right away. We spent like 3 weeks fixing boring hooks. Once that got better, cutting podcast clips stopped eating the whole afternoon.
UK · newsletter operator
generated scenario
I thought it would be a shorts thing, but it became a newsletter repurposing thing. Research in, LinkedIn clips and archive videos out.
Canada · B2B consultant
generated scenario
For consulting it was not plug and play. Had to build examples first. After that, the short explainers made sales calls way less repetitive.
India · fitness programmer
generated scenario
This ended up being more useful for members than for viral posts. Weekly plans turned into app updates, email clips, and a few decent acquisition videos.
Because the base repo and seven shipped modules share one plugin contract. You save when you buy the full surface together, and setup stays focused on channel decisions instead of repeated checkout and dependency sequencing.
The buyer repo now presents AI mode and Manual mode clearly, includes clip-farming docs, ships TikTok publishing as a real module surface, and exposes Analytics, Clip Harvester, Viral Captions, and AI Rulesets through the plugin loader.
Yes, but pick the right entry point. The Concept PDF gives the system map, the Manual-mode repo path avoids LLM spend, and the Full AI pipeline is for buyers ready to configure provider keys and operate the repo.
// choose the next artifact
Get the source-code workflow, docs, and setup path for publishing Shorts with your own accounts, provider keys, and module stack.