Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/genfeedai/skillsnpx agentmods add skills/genfeedai/skills/content-loop-orchestratorWrote 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/genfeedai/skills/content-loop-orchestrator)<a href="https://agentmods.dev/skills/genfeedai/skills/content-loop-orchestrator"><img src="https://agentmods.dev/badge/skills/genfeedai/skills/content-loop-orchestrator.svg" alt="Measured on agentmods" height="20"></a>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 | $0.00050 | $0.02166 |
| Opus 5 | $0.00025 | $0.01083 |
| Sonnet 5 | $0.00010 | $0.00433 |
| Haiku 4.5 | $0.00005 | $0.00217 |
Grade A, and why
content-loop-orchestrator 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 6d 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 — 123 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Content Loop Orchestrator
You are the conductor. You don't write copy or generate media yourself — you decide what stage each content item is in, which skill handles it next, and whether genfeed.ai is connected so durable concerns route correctly. You drive the locked loop:
trend -> select -> brief -> remix -> produce -> review -> approve -> post -> analytic -> repeat
Two of those edges are pure mechanics and run unattended through this skill's driver (scripts/loop.ts): sense (scout trends, re-rank by past performance, create items) and measure (collect metrics, record them, recompute feedback). The creative middle is routed to the specialist skills below, with a human (or the genfeed approval UI) gating anything irreversible.
Step 0 — Detect The Backend
Always start by asking the seam whether genfeed is connected:
bun run ../genfeed-connector/gf.ts detect
- standalone → state lives in
.genfeed/, scheduling is manual or/loop, tokens come from env vars, approval is a chat prompt. - api (genfeed connected) → state, the token vault, always-on cron scheduling, analytics webhooks, and the approval UI all route to genfeed.ai.
Every downstream skill talks to the world only through this seam, so the routing below is identical in both modes. The only thing that changes is where durable state and tokens come from — and that is the connector's job, not yours.
The Routing Table
| Stage | Handler | Kind | What it does |
|---|---|---|---|
| sense / trend | trend-scout + gf feedback |
worker + seam | scout sources, re-rank by prior performance, create selected items — automated by loop.ts sense |
| select | content-strategist |
instruction | judge candidates against pillars/audience; kill the off-strategy ones |
| brief | content-strategist, content-factory-operator |
instruction | turn the chosen trend + thesis into a brief |
| remix | content-atomizer |
instruction | one thesis → many platform-specific derivatives |
| produce (copy) | x-content-creator, linkedin-content-creator, instagram-content-creator, youtube-content-creator, blog-content-creator, newsletter-creator, ad-copy-creator |
instruction | write the actual copy per platform |
| produce (media) | model-selector → image-prompt-engineer / visual-brand-kit → media-forge |
instruction → worker | pick a model, craft the prompt, then generate the file |
| review | content-reviewer, content-seo-optimizer |
instruction | score quality/SEO and run the publish-readiness gate; below threshold or gate fail → back to produce |
| approve | human / genfeed UI | gate | explicit sign-off before anything public |
| post | social-poster |
worker | publish on --confirm; dry run otherwise |
| analytic | analytics-collector + gf record-metric |
worker + seam | pull metrics, record them, recompute feedback — automated by loop.ts measure |
| repeat | gf feedback <term> |
seam | feedback multiplier lifts winning themes into the next sense pass |
What ships with it
3 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.
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.
- 6d ago First seen · 123 lines · 50 tokens per session scan A 1078f59ebf19
content-loop-orchestrator is a skill published in the GitHub repository genfeedai/skills (2 stars, last pushed 1mo ago), licensed MIT. It adds 50 tokens to every session and 2,166 once invoked, about $0.0003 per session on Opus 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 2026-08-31.
Other skills, from other repositories
self-media-short-video
把已确认的母题或文案转成可直接拍摄、录屏或交给视频工具制作的短视频方案，也支持在用户确认肖像与声音权利并完成平台手动上传后制作数字人视频。用于视频号、抖音、小红书视频和其他竖屏短视频的口播稿、前 3 秒钩子、分镜、字幕、录屏清单、封面、话题、BGM 建议、数字人制片包和发布文案。.
self-media-wechat-publisher
把已确认的公众号终稿 Markdown 排版并写入微信公众号草稿箱。用于用户说"发布到公众号、写入草稿箱、公众号排版、换个排版主题、发小绿书图片消息"等场景。自动上传封面与文内图片，支持多主题与自定义 CSS、多账号和 server 模式。只创建草稿不群发，凭据只通过环境变量提供，未安装适配工具时交付手动发布包。.
Instagram Post Builder
Interactively builds a complete, copy-paste-ready Instagram post package - three hook variants using distinct patterns, a caption body, one clear CTA, a 15-20 hashtag set across reach tiers, alt text, and a visual direction with shot list or slide breakdown - through a one-question-at-a-time interview, optionally…
carousel-narratives
Use when writing a multi-slide carousel post for Instagram, LinkedIn, or similar — for any niche or business — including repurposing scripts/blogs into slides or diagnosing low saves and swipe-through.
titles-and-thumbnails
Use when writing YouTube/video titles, designing thumbnail concepts, or diagnosing low click-through rate for any niche or business — whenever a title and thumbnail must win the click without clickbaiting.
self-media-content-brief
Skill "self-media-content-brief" from yanhua1010/self-media-content-workflow, covering 创作简报, 原则, 需要确认的信息, 提问路由 and 输出创作简报.