I
inreels-ai/inreels-skills
Open agent skills for making ads — UGC video ads, static ad creatives and faceless shorts — from Claude,…
在 GitHub 查看 ↗链接已于 2026-11-12 核查 · HTTP 200 · 打开的是厂商自己的官网
- ★ 数取值
- 2 ★ · inreels-ai/inreels-skills · 取自 GitHub 搜索 API · (2026-11-13 20:00:50 核对)
- 来源明细
- 从该仓库的 GitHub 页面直接读取 · 条目类型:开源项目
- 收录与活跃
- 收录于 2026-11-08 · 最近见到 2026-11-14 · 共见到 7 次 · 被 1 个来源列出
- 评分构成
- 综合分 103 · 2 ★ · 1 个来源 · repo
常见问题(答案都来自我们的记录)
它适合谁 / 用来做什么?
我们把它归在这些标签下: 开源项目。标签来自条目自身,不是我们强加的分类。
谁维护它、还在更新吗?
仓库
inreels-ai/inreels-skills;我们收录于 2026-11-08;最近见到 2026-11-14。这里没有"官方认证"或"质量评级"——我们只记录来源与时间。# inReels Agent Skills
以下摘录自该仓库的 README(片段) · 在 GitHub 读完整 README →
Open skills for making ads, UGC video and faceless shorts from an AI agent,
built on the inReels MCP server.
The skills are MIT licensed and free. The rendering service they call is paid.
## Install
Point your agent at the install instructions and let it do the work:
Install the inReels skills: https://raw.githubusercontent.com/inreels-ai/inreels-skills/main/install.md
Or by hand:
bash
# 1. Add the MCP server
claude mcp add --transport http --scope user inreels https://www.inreels.ai/api/mcp
# 2. Copy the skills you want
git clone https://github.com/inreels-ai/inreels-skills
cp -r inreels-skills/skills/ ~/.claude/skills/
Then run /mcp, pick inreels and sign in from the browser. Or create a key
at and set it as the
server's bearer token.
Cursor, Codex, ChatGPT and Muse are all supported —
see .
## The skills
Make something
| Skill | What it does |
| animated-short | One line → an animated short film: story, consistent character, narration, sound, rendered MP4 |
| claymation-ad | A stop-motion clay commercial: clay characters intercut with photoreal macro, cut to a real broadcast beat map |
| ugc-ad | Product image + creator + spoken script → rendered UGC video ad, with native audio |
| product-demo-ad | No presenter — the product doing its job, in motion |
| ad-creative | Angle → hook → rendered static ad creative, sized for the placement |
| faceless-video | Retention-shaped script → scenes, narration, captions → rendered MP4 |
| scary-story-video | Horror, true crime and unsolved mystery shorts, and the craft they need |
Craft the others lean on
| Skill | What it does |
| hook-writing | The first two seconds — five shapes that work, and testing them properly |
| ad-angle-testing | A set of ads that disagree usefully, instead of one ad four times |
| character-consistency | Keep the same person or product across several generated clips |
Each one states its method, names the failure modes it has actually seen, and
is honest about what it does not do.
animated-short and claymation-ad are the odd ones out: they ship working scripts and run on
whatever backend you point it at — Replicate, fal, ElevenLabs, or the inReels
MCP as the no-key option. The rest are craft and method for the MCP server
below.
## What the MCP server exposes
17 tools across two layers.
Capabilities return a finished deliverable and choose the model for you:
create_ad_video — a finished ad from a product image, with native audio.
Add a creator and it becomes a UGC ad.
create_faceless_video — script to scenes to narration to captions to MP4.
quote_ad_video, quote_faceless_video — exact price first.
Primitives render one asset and expose the full parameter surface:
generate_image, generate_video, generate_speech
quote — exact price of any primitive call, charging nothing.
Discovery and plumbing, all free:
list_models, list_voices, list_styles, list_actors
get_run_status, list_runs, get_account
upload_image, refresh_asset_url
## Conventions worth knowing
built on the inReels MCP server.
The skills are MIT licensed and free. The rendering service they call is paid.
## Install
Point your agent at the install instructions and let it do the work:
Install the inReels skills: https://raw.githubusercontent.com/inreels-ai/inreels-skills/main/install.md
Or by hand:
bash
# 1. Add the MCP server
claude mcp add --transport http --scope user inreels https://www.inreels.ai/api/mcp
# 2. Copy the skills you want
git clone https://github.com/inreels-ai/inreels-skills
cp -r inreels-skills/skills/ ~/.claude/skills/
Then run /mcp, pick inreels and sign in from the browser. Or create a key
at and set it as the
server's bearer token.
Cursor, Codex, ChatGPT and Muse are all supported —
see .
## The skills
Make something
| Skill | What it does |
| animated-short | One line → an animated short film: story, consistent character, narration, sound, rendered MP4 |
| claymation-ad | A stop-motion clay commercial: clay characters intercut with photoreal macro, cut to a real broadcast beat map |
| ugc-ad | Product image + creator + spoken script → rendered UGC video ad, with native audio |
| product-demo-ad | No presenter — the product doing its job, in motion |
| ad-creative | Angle → hook → rendered static ad creative, sized for the placement |
| faceless-video | Retention-shaped script → scenes, narration, captions → rendered MP4 |
| scary-story-video | Horror, true crime and unsolved mystery shorts, and the craft they need |
Craft the others lean on
| Skill | What it does |
| hook-writing | The first two seconds — five shapes that work, and testing them properly |
| ad-angle-testing | A set of ads that disagree usefully, instead of one ad four times |
| character-consistency | Keep the same person or product across several generated clips |
Each one states its method, names the failure modes it has actually seen, and
is honest about what it does not do.
animated-short and claymation-ad are the odd ones out: they ship working scripts and run on
whatever backend you point it at — Replicate, fal, ElevenLabs, or the inReels
MCP as the no-key option. The rest are craft and method for the MCP server
below.
## What the MCP server exposes
17 tools across two layers.
Capabilities return a finished deliverable and choose the model for you:
create_ad_video — a finished ad from a product image, with native audio.
Add a creator and it becomes a UGC ad.
create_faceless_video — script to scenes to narration to captions to MP4.
quote_ad_video, quote_faceless_video — exact price first.
Primitives render one asset and expose the full parameter surface:
generate_image, generate_video, generate_speech
quote — exact price of any primitive call, charging nothing.
Discovery and plumbing, all free:
list_models, list_voices, list_styles, list_actors
get_run_status, list_runs, get_account
upload_image, refresh_asset_url
## Conventions worth knowing
收录徽章(可选)
想把收录标记放进 README 就复制这行,可选,不影响收录。
[](https://anclum.com/zh/s/inreels-ai-inreels-skills/)
- 来源
- github · https://github.com/inreels-ai/inreels-skills
- 按领域浏览
- 开源项目
- Traction
- 2 ★
- 收录于
- 今天
同类推荐
所属领域与标签
每个标签都能点进去看同类的全部条目。数字是当前库里的真实条数。