trypostit/trypost

Open-source social media scheduler built for AI agents and the teams behind them. Create, schedule and publish posts to every network through MCP or the REST API, or write them yourself in a composer your team will love.

TryPost – Open‑source social‑media scheduler for AI agents

What it is — TryPost is a self‑hostable web application that lets you create, schedule and publish posts across dozens of social networks. It is designed for two audiences:

  • AI agents – it exposes a MCP server (with 80+ built‑in tools) and a conventional REST API so large language model assistants such as Claude, ChatGPT or custom bots can program‑matically generate content, queue it, and pull analytics.
  • Human teams – a rich web UI with a composer, calendar view, approval workflow, workspaces, and an “ideas board” for editorial planning.

Key features

Area What you get
Automation interface MCP server + REST API; same capabilities as the UI (post creation, media upload, queue management, analytics, webhooks).
Composer & editor Write once, adapt per network; live preview; AI‑assistant for re‑phrasing, shortening or expanding text.
Scheduling Calendar view, per‑channel queues, drag‑and‑drop rescheduling, automatic slot selection.
Collaboration Approvals, comments/notes with email notifications, reusable signatures/labels, multi‑workspace isolation.
Analytics Followers, reach, engagement per channel and per post, with period‑over‑period comparison.
Cross‑posting Automatic repurposing of media to other connected networks.
Media sources Direct upload, Unsplash, Google Drive/Photos, Canva.
Integrations Native posting to 13+ networks (Instagram, TikTok, LinkedIn, X, YouTube, Facebook, Threads, Bluesky, Pinterest, Mastodon, Google Business, Telegram, Discord).
Extensibility Webhooks for lifecycle events, 16‑language UI.

Why it matters for AI

  • The platform treats the AI‑generated workflow as first‑class: the same permission model, queue limits, and approval steps apply to both bots and humans.
  • By offering a MCP (Message‑Control‑Protocol) server with a large toolbox, developers can plug LLM‑driven agents directly into a production‑grade scheduler without writing custom glue code for each network.
  • The open‑source AGPL‑3.0 license lets organisations self‑host, keeping proprietary content and analytics on‑premise—a common requirement for enterprises that let internal LLMs manage brand communication.

Typical use‑case

  1. An LLM agent receives a marketing brief (e.g., “promote spring tours”).
  2. It calls the TryPost MCP API to generate a week’s worth of captions, selects appropriate hashtags from a stored signature, and uploads any promotional images.
  3. The agent queues the posts for Instagram, LinkedIn and X.
  4. A human reviewer approves the batch via the web UI; once approved, TryPost publishes automatically at the pre‑configured times.
  5. Post‑publish, the agent queries the analytics endpoint to retrieve reach/engagement numbers for the last month.

Getting started

  • Self‑host – clone the repo, follow the documentation (Docker compose is typical) and run the AGPL‑licensed server on your own infrastructure.
  • Try the SaaS – sign up at trypost.it for a hosted version that mirrors the open‑source feature set.
  • Extend – use the REST API or MCP tools to write custom scripts, integrate with your own LLM pipelines, or add new network connectors.

Community – Discord, GitHub Discussions, and an open contribution model (issues, PRs). The project is maintained by Paulo Castellano and welcomes both code contributions and feedback.

All information above is taken directly from the repository’s README.

Related

  • Project
    plasma-ai/fractal

    fractal is a Python library/CLI that runs autonomous LLM agents in a recursive, tree‑structured set of Git worktrees. Nodes iterate toward goals, can spawn child nodes for sub‑tasks, and are bounded by configurable caps on depth, iterations, time, and USD cost. All run metadata is stored in a local SQLite DB and visualised via a terminal UI. Supports Claude Code, Codex, Grok, OpenCode, and Oh My Pi back‑ends, with a `/fractal` skill for chat‑based node creation. Safety warning: nodes run without permission prompts, so they should be launched only on trusted or sandboxed hosts.

  • Project
    ljquan/opentu

    Opentu is a canvas-based AI application platform that integrates multi-model generation and tool management into a single workspace for continuous AI task execution.

  • Project
    auscaster/frantic-board

    A public GitHub board for the Frantic platform, where AI agents can claim paid bounties, deliver work, and receive USDC. The experiment tracks how well autonomous or human‑assisted agents perform real commercial tasks, using the runx runtime for verifiable execution receipts.

  • Project
    dualface/ste-zh

    ste‑zh is an open‑source skill that makes LLM agents write their Chinese status reports in a strict, evidence‑based format inspired by the aerospace standard ASD‑STE100. It can be installed via the `skills` CLI or manually, activated with simple commands, and is useful for teams that need clear, concise AI‑generated updates.