Posta is for developers, AI agents, automation engineers, and the creators they support — anyone who treats social posting as a pipeline rather than a manual task. Plug Posta into an MCP server, an n8n workflow, a Claude Code skill, or the REST API and let the agent do the rest.

What is the Posta MCP server?

Posta ships an MCP (Model Context Protocol) server so any MCP-capable client — Claude Desktop, Claude Code, Cursor, or your own agent — can create, schedule, and publish social media posts as tool calls. One install (npx posta-mcp), one token, every supported network.

Does Posta work with n8n?

Yes. The n8n-nodes-posta community node turns every Posta endpoint into a typed n8n node. We publish ready-to-fork workflows for blog-to-social, YouTube-to-social, product launch campaigns, and LinkedIn carousel generation — wire any trigger to Posta in a few clicks.

Can I post to social media from Claude Code or the terminal?

Yes. The Posta Claude Code skill lets you create, schedule, and publish posts straight from your terminal or IDE using natural-language slash-commands. No dashboard required, no context-switch from your editor.

Does Posta publish directly to social networks?

Yes. Posta publishes directly to LinkedIn, TikTok, Instagram, YouTube, Pinterest, Facebook, Bluesky, and Threads via each platform's official API, and fires HMAC-signed webhooks back to your agent when posts go live — so closed-loop pipelines know exactly when they can act on a result.

How is this different from Buffer, Hootsuite, or Later?

Those tools are dashboards designed for humans clicking buttons. Posta is API-first and agent-first: full public REST API on every paid tier, HMAC-signed outbound webhooks, an MCP server, a Claude Code skill, and an n8n community node — none of which the dashboard-led incumbents ship.

Can I influence the roadmap?

Yes. Early users help shape integrations, workflow and pricing. Tell us how you work and what your current agent stack looks like when you request access.

Post to Bluesky from the terminal with Claude Code

Bluesky's native app and web client are fine, but if you live in a terminal anyway, switching to draft a 280-char post is friction. The Posta Claude Code skill lets you draft and post Bluesky entries with the same agent that's editing your code — natural-language prompts, no slash-commands, no context switch.

Setup

You need a Posta account, a connected Bluesky account, and Claude Code. If you have all three, this is the whole install — one command in Claude Code, pointing at the skill repo:

# In Claude Code:
/install-skill https://github.com/STGime/posta-skill

# Once installed, Claude can call Posta in natural language —
# no Posta-specific slash-commands required.

Connect Bluesky from the Posta dashboard's Accounts page first; the skill picks up your connected account automatically. The full reference is on the CLI posting landing page.

Pattern 1 — Single Bluesky post under 300 chars

Bluesky's post limit is 300 graphemes. Ask Claude to draft and post under the cap:

"Draft a Bluesky post under 300 chars about our v2 launch.
Be specific — mention per-platform caption limits and the batch
media endpoint. Schedule for now (so I can publish from the app
if I like it). Save as a draft."

Claude calls Posta to create the draft on Bluesky. Posta enforces Bluesky's 300-char limit on its side — if the draft is too long, the response surfaces the overage so Claude can rewrite tighter.

Pattern 2 — Bluesky thread from a repo

Threads work well on Bluesky for technical commentary. Ask Claude to draft a multi-post thread from a recent commit or PR:

"Look at PR #43 (the SEO Phase 3 batch 3 PR). Draft a 4-post
Bluesky thread about what shipped: the Vercel AI SDK tutorial,
the Mastra tutorial, and the Zapier-migration piece. Keep each
post under 300 chars. Schedule the whole thread for tomorrow 9am
CET as drafts so I can review."

Claude reads the PR, picks the thread shape, drafts each post, and calls Posta for the whole sequence. Bluesky threads are first-class in Posta — the skill knows to chain them.

Pattern 3 — Cross-post from a longer LinkedIn draft

Already drafted something long-form for LinkedIn? Ask Claude to compress for Bluesky:

"Take my last LinkedIn draft (the one about the v2 launch),
compress to a single Bluesky post under 300 chars, and save as
a Bluesky draft. Keep the technical specifics — drop the
audience-warming opener."

Why the skill, not the MCP server?

You can also install the Posta MCP server into Claude Code. The skill is the focused wrapper around the common posting workflows (draft, schedule, list, publish), tuned for Claude Code specifically. The MCP server exposes Posta's full tool surface — createPost, schedulePost, listAccounts, getPostStatus, listMedia, uploadMedia, listPosts — to any MCP client. Most users start with the skill and add the MCP server when they need to call the broader tool set from Claude Desktop, Cursor, or a custom agent.

Where to go from here

Same skill works for LinkedIn — see post to LinkedIn from the terminal with Claude Code. For the broader decision framework on which surface to pick, see MCP vs n8n vs Claude Code. 14-day free trial covers Bluesky plus seven other networks.