填补了Substack缺乏原生发布API的空白，让作者能通过AI代理实现内容工作流自动化。
Substack does not provide a native public API for publishing posts.
But authors who want to draft, publish, and schedule from AI agents (Cursor, Claude, and other MCP hosts) still need a way out. This is the MCP server to publish to Substack via any agent.
Will not start without a valid APISUBSTACK_API_KEY (ask_*). Get API key here: apisubstack.com.
| Reality | What we built |
|---|---|
| No official Substack write/publish API for authors | A posting MCP so agents can create drafts, publish now, and schedule |
| Authors live in the browser editor | Authors publish through their agents instead |
| Session-cookie auth is the practical path today | substack.sid + your publication URL, gated by an API Substack key |
| Tool | Description |
|---|---|
test_connection | Verify substack.sid and return profile |
create_post | Create draft / publish now / schedule (+ optional tags) |
get_draft | Fetch draft by id |
update_draft | Update draft fields |
publish_post | Publish existing draft |
schedule_post | Schedule existing draft |
delete_draft | Delete draft |
list_tags | List publication tags |
create_tag | Create tag |
set_tags | Ensure + attach tags |
get_post_tags | List tags on a post/draft |
cd /path/to/Substack-API-MCP
python3 -m venv .venv
source .venv/bin/activate
pip install -e .
ask_* API key.substack.sid cookie:
https://substack.com (or your publication host, e.g. https://yourname.substack.com).substack.sid and copy its Value.export APISUBSTACK_API_KEY="ask_YOUR_KEY"
export SUBSTACK_PUBLICATION_URL="https://yourname.substack.com"
export SUBSTACK_SID="YOUR_SUBSTACK_SID_VALUE"
# optional
# export SUBSTACK_USER_ID="123456"
Cookies expire. If you get 401/403, refresh substack.sid and try again.
substack-api-mcp
# or
python -m substack_api_mcp
See `mcp.example.
MCP server for Substack posts (FastMCP + substack-api-client)
根据分类、Topic 和编程语言匹配的相似项目。
Deltafin is a research project that runs the 2.8-trillion-parameter Mixture-of-Experts model Kimi K3 on a single Apple Silicon Mac (e.g., M1 Max with 64 GB) at about 16 seconds per token, using exact, reproducible inference with local or streaming expert loading.
Clodex is an open-source, local-first agentic IDE that combines persistent AI tasks, code, terminal, browser, Git, models, memory, and governed execution in one Electron workspace, currently in technical preview.
OpenClaude is an open-source CLI coding agent that runs on any platform and supports a wide range of LLM providers, offering the same tools and workflows as Claude Code.