Dropspace Content Engine
Fully autonomous content pipeline. Every night: pull analytics → identify winning hooks → generate new posts → schedule across 6 platforms. The feedback loop compounds — each cycle produces better content because it learns from real engagement data.
Setup
1. Clone and install
CODEBLOCK0
If canvas fails to install, that's fine — text-only formats still work. For visual formats (TikTok/Instagram slideshows): macOS brew install pkg-config cairo pango, Linux apt install libcairo2-dev libpango1.0-dev, then npm install again.
2. Run the setup wizard
CODEBLOCK1
Walks you through API keys, platform selection, and app configuration.
3. Set your API keys
CODEBLOCK2
Save in a .env file (copy from templates/.env.example). Add .env to .gitignore to avoid committing secrets.
4. Validate
CODEBLOCK3
Run the Pipeline
CODEBLOCK4
Automate Nightly
| Time (ET) | Script | What |
|---|
| 12:00 AM | INLINECODE8 + INLINECODE9 | Refresh analytics, clean old media |
| 12:30 AM |
scripts/run-x-research.js --app myapp | Scan X for trending hooks (optional, needs Bird CLI) |
| 1:00 AM |
scripts/run-self-improve-all.js --app myapp | Analyze + generate (up to 60 min) |
| 2:00 AM |
scripts/schedule-day.js --app myapp | Schedule for today |
What Happens Each Run
- 1. Pulls 14 days of analytics from Dropspace API
- Identifies winning hooks (posts with >2x average engagement)
- Cross-references with X/Twitter research signals
- Generates 7-14 new posts with AI (text, visual, video formats)
- Fact-checks any claims about people, products, or events
- Writes strategy notes that persist between runs (the loop compounds)
- Schedules posts across the day via Dropspace API
Content Formats
| Format | Type | Platforms |
|---|
| story-slideshow | visual | TikTok, Instagram, Facebook |
| ugc-reaction |
video | TikTok, Instagram |
| text-single | text | Twitter/X |
| text-post | text | LinkedIn, Reddit, Facebook |
Links
- - Community page: https://www.dropspace.dev/community/dropspace-content-engine
- Case study: https://www.dropspace.dev/case-studies/march-2026
- Repo: https://github.com/joshchoi4881/dropspace-agents
- Dropspace API docs: https://www.dropspace.dev/docs
Dropspace 内容引擎
完全自主的内容流水线。每晚：拉取分析数据 → 识别爆款钩子 → 生成新帖子 → 在6个平台安排发布。反馈循环持续累积——每一轮都能产出更好的内容，因为它能从真实的互动数据中学习。
设置
1. 克隆并安装
bash
git clone https://github.com/joshchoi4881/dropspace-agents && cd dropspace-agents && npm install
如果 canvas 安装失败，没关系——纯文本格式仍然可用。对于视觉格式（TikTok/Instagram 幻灯片）：macOS 运行 brew install pkg-config cairo pango，Linux 运行 apt install libcairo2-dev libpango1.0-dev，然后再次运行 npm install。
2. 运行设置向导
bash
node setup.js --template dropspace-content-engine
引导你完成 API 密钥、平台选择和应用配置。
3. 设置你的 API 密钥
bash
export DROPSPACEAPIKEY=dslive... # 来自 dropspace.dev/settings/api
export ANTHROPICAPIKEY=sk-ant-... # 来自 console.anthropic.com
export FALKEY=fal... # 来自 fal.ai（用于视觉/视频格式）
保存在 .env 文件中（从 templates/.env.example 复制）。将 .env 添加到 .gitignore 以避免提交密钥。
4. 验证
bash
node scripts/test-pipeline.js --app myapp
运行流水线
bash
source .env
分析表现 + 生成新帖子
node scripts/run-self-improve-all.js --app myapp
安排今天生成的帖子
node scripts/schedule-day.js --app myapp
每晚自动化
| 时间（美东时间） | 脚本 | 操作 |
|---|
| 凌晨12:00 | scripts/refresh-tracking.js --all + scripts/cleanup-posts.js --all | 刷新分析数据，清理旧媒体 |
| 凌晨12:30 |
scripts/run-x-research.js --app myapp | 扫描 X 平台的热门钩子（可选，需要 Bird CLI） |
| 凌晨1:00 | scripts/run-self-improve-all.js --app myapp | 分析 + 生成（最多60分钟） |
| 凌晨2:00 | scripts/schedule-day.js --app myapp | 安排今天的帖子 |
每次运行的内容
- 1. 从 Dropspace API 拉取14天的分析数据
- 识别爆款钩子（互动率超过平均水平2倍的帖子）
- 与 X/Twitter 研究信号交叉引用
- 使用 AI 生成7-14篇新帖子（文本、视觉、视频格式）
- 对涉及人物、产品或事件的任何声明进行事实核查
- 编写在每次运行之间持续存在的策略笔记（循环累积）
- 通过 Dropspace API 在一天内安排帖子发布
内容格式
| 格式 | 类型 | 平台 |
|---|
| story-slideshow | 视觉 | TikTok, Instagram, Facebook |
| ugc-reaction |
视频 | TikTok, Instagram |
| text-single | 文本 | Twitter/X |
| text-post | 文本 | LinkedIn, Reddit, Facebook |
链接
- - 社区页面：https://www.dropspace.dev/community/dropspace-content-engine
- 案例研究：https://www.dropspace.dev/case-studies/march-2026
- 仓库：https://github.com/joshchoi4881/dropspace-agents
- Dropspace API 文档：https://www.dropspace.dev/docs
