Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add realjaymes/marketingagentskills --skill ai-content-generationgit clone --depth 1 https://github.com/realjaymes/marketingagentskillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/realjaymes/marketingagentskills/ai-content-generation)<a href="https://agentmods.dev/skills/realjaymes/marketingagentskills/ai-content-generation"><img src="https://agentmods.dev/badge/skills/realjaymes/marketingagentskills/ai-content-generation/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/realjaymes/marketingagentskills/ai-content-generation"><img src="https://agentmods.dev/badge/skills/realjaymes/marketingagentskills/ai-content-generation.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00243 | $0.02634 |
| Opus 5.5 | $0.00097 | $0.01054 |
| Sonnet 5.5 | $0.00049 | $0.00527 |
| Haiku 4.5 | $0.00024 | $0.00263 |
Grade A, and why
ai-content-generation scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 15d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 89 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AI Content Generation
Help the user make studio-quality AI video and image content with the fewest steps. Every use case runs the same six-step workflow. A proven reference does most of the planning, the script and voice set the pacing, and the user approves the script and the start frames before any credits go into video.
The full workflow and the shared rules are in references/00-workflow-and-rules.md. Read it before the first step of any job.
Pick the playbook
Ask first: is there a proven reference? If not, find the story first with the four stops in the Story craft rules (references/00-workflow-and-rules.md), then find a reference for the style. Then route:
| The user wants to... | On screen | Reference file |
|---|---|---|
| Run a recurring AI character on TikTok or Instagram | An AI person (type B) | references/01-faceless-influencer.md |
| Clone themselves and generate talking-head videos | Their clone (type B) | references/02-ai-clone-talking-head.md |
| Make paid ad creative at volume, including AI UGC | An AI person or product (type B) | references/03-ai-ad-creative.md |
| Build a faceless YouTube channel | B-roll under narration (type A) | references/04-faceless-youtube.md |
| Make a product or launch video with motion graphics, with an AI voice or none | No person (type A) | references/05-launch-videos-and-recording-edits.md |
| Make 15 to 25 second demos of a product feature or free tool (tool shorts, twelve styles) | No person (type A) | references/05-launch-videos-and-recording-edits.md |
| Edit their own recording, or copy a creator's edit style | Themselves (type C) | references/05-launch-videos-and-recording-edits.md |
"Faceless" means the operator's own face is absent. An AI character in a faceless account has a face. Never pass a generated or stock person off as a real, named individual.
The six steps
- Reference brief. A five-line brief plus one to three proven references, torn down with
references/reference-teardown.md. To copy a style, measure it in numbers. A style we will reuse becomes a Remotion template. - Script. A beat table with a word budget per beat that follows the reference's story. Every claim traces to the brief; flag missing facts and never invent them. Every script has a hook, a pain, a solution and a CTA, told through one launch story angle or a character story built by the Story craft rules. Every line is something a person actually says, in short lines, with no calendar markers or odd counts as fake specificity, and each beat causes the next. Codex drafts or reviews every script. A video that sells a program or offer runs 90 to 120 seconds and is built from the offer's landing page (
references/03-ai-ad-creative.md, story and sale videos). Study proven references closely and get a Codex second opinion before writing. The user approves the script. A story script goes to the user with its storyboard, and both are approved together. - Scenes and anchors. Character, product and look anchors, then one row per shot with its full spec. A story script's storyboard (one code-drawn frame per shot, no image generation, in the same brief file) is approved before any face or start frame is made. Generate one start frame per generated shot and show them as a contact sheet. The user approves the start frames.
- Voice. The final voice comes first. Whisper times every word into
words.json, and that sets each shot's length. A coded launch film that gets a voice after its picture is approved is the exception: the picture stays fixed and one short line per beat is fitted to it. - Shots. One take per shot from its start frame in Google Flow, with the settings and per-shot model choice in the Google Flow rules. Fix faults in the edit first. At most one retry per clip, with the user's go-ahead.
- Finish and check. Edit generated clips by the clip-editing rules (never joined end to end), add sound by the video type's rule and word-timed captions, match colour, set the platform AI label, run the final check, and make variants for ads.
What ships with it
16 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- assets/chatgpt-project/knowledge/00-workflow-and-rules.md 44 KB
- assets/chatgpt-project/knowledge/01-faceless-influencer.md 20 KB
- assets/chatgpt-project/knowledge/02-ai-clone-talking-head.md 27 KB
- assets/chatgpt-project/knowledge/03-ai-ad-creative.md 42 KB
- assets/chatgpt-project/knowledge/04-faceless-youtube.md 26 KB
- assets/chatgpt-project/knowledge/05-launch-videos-and-recording-edits.md 41 KB
- assets/chatgpt-project/knowledge/reference-teardown.md 5.0 KB
- assets/chatgpt-project/project-instructions.md 3.8 KB
- evals/evals.json 3.9 KB
- references/00-workflow-and-rules.md 44 KB
- references/01-faceless-influencer.md 19 KB
- references/02-ai-clone-talking-head.md 26 KB
- references/03-ai-ad-creative.md 42 KB
- references/04-faceless-youtube.md 26 KB
- references/05-launch-videos-and-recording-edits.md 41 KB
- references/reference-teardown.md 4.9 KB
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 15d ago Changed · +5 lines · -1 tokens per session b90bda818451
- 18d ago Changed · -42 lines · +58 tokens per session e338248938f7
- 1mo ago First seen · 126 lines · 186 tokens per session scan A a910950190c1
ai-content-generation is a skill published in the GitHub repository realjaymes/marketingagentskills (68 stars, last pushed 2d ago), licensed MIT. It adds 243 tokens to every session and 2,634 once invoked, about $0.0010 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
local-ai-agents
Build local-first AI agents that run entirely on a developer workstation with Microsoft Foundry Local and Qwen function-calling models. Covers Small Language Models (SLMs), the OpenAI-compatible local endpoint, sandboxed local tools, local RAG with Chroma, local MCP servers, hybrid cloud/local routing, and the…
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
next-cache-components-adoption
Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…
chat-pet-sprite-creation
Use when creating or changing VS Code chat pet sprite art, sprite sheets, state animations, eye treatments, Stable/Insiders variants, or pet transitions under src/vs/workbench/contrib/chat/browser/widget/media/chatPet.
cpu-profile-analysis
Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…
next-cache-components-optimizer
Optimize the meaningful UI available immediately from a Next.js route on an initial load (hard navigation) or named client-side navigation (soft navigation). Encode the goal as a failing @next/playwright instant() e2e and work it to green, one verified route and entry point at a time; the shipped test then guards…