pendpost
pendpost is a free, open-source social media planner that uses an AI agent to draft and schedule posts across Instagram, Facebook, LinkedIn, YouTube, X, Telegram, Discord, Mastodon, Nostr, WordPress, Ghost, Reddit, Pinterest, TikTok, and Google Business Profile.
Ficha actualizada 23 sept 2026, 14:07 · upstream 23 sept 2026, 13:57 · release v2.6.1
Qué es pendpost
Descripción editorial construida a partir del repositorio, su documentación y sus señales de mantenimiento.
pendpost is an agent-first social media platform where an AI agent drafts and schedules posts across multiple platforms behind a human approval gate you control. Unlike traditional schedulers that simply queue content, pendpost treats the approval process as a core part of its workflow — every post carries an approval state (draft, approved, rejected) and only moves forward once it passes your review. This design prevents accidental publishing while still giving you full agency over what gets published.
The tool is built on top of MCP (Model Context Protocol), meaning any agent that supports MCP can integrate with pendpost by simply connecting to the local server at http://127.0.0.1:8090/mcp. This makes it easy for developers and AI agents to consume the platform as a service without needing to manage their own servers or credentials.
Under the hood, pendpost uses Node.js with React for the web dashboard and handles scheduling via native APIs where available (Facebook scheduled posts, YouTube publishAt) while falling back to queue-based scheduling for platforms that don't support it. It also includes a health-check endpoint at /health that returns status information about all connected clients.
- A developer team needs to plan and schedule content across multiple social channels without manually creating posts each day
- An agency wants to use AI agents to draft and approve content while maintaining human oversight for brand safety
- A small business owner runs a local server with pendpost and uses it as their primary social media management tool
- A developer is building an MCP client that integrates with pendpost to automate posting workflows across multiple platforms
- An agency wants to offer AI-assisted content creation services while keeping human approval for final publishing decisions
- A team needs a local-first solution that doesn't require cloud infrastructure and can run entirely offline
- pendpost is completely free, open-source under the MIT license with no hidden costs or subscription requirements
- It runs locally on any machine without needing Docker or external servers — everything executes in-memory via Node.js
- The platform supports MCP (Model Context Protocol) natively, allowing any agent to integrate with it through a simple HTTP endpoint
- Every post carries an approval state that prevents accidental publishing and gives you full control over what goes live
- It includes built-in brand-lint rules and humanizer checks that catch common AI writing mistakes before posts are published
- The tool supports both web-based dashboards and MCP clients, making it accessible to developers, agencies, and non-developers alike
- pendpost is not designed for high-volume posting — it is a planner and scheduler, not an auto-posting service that can handle thousands of posts per day
- The approval gate means you cannot publish content automatically without human intervention; this adds friction to workflows that require speed
- Some platforms like X (Twitter) do not support native scheduling via their APIs, so pendpost must queue those posts until the platform supports it
- The tool requires a local Node.js environment running on each machine where you want to publish — there is no cloud-based or serverless deployment option
- While MCP-native, the tool does not provide any built-in authentication or credential management — all credentials must be stored in your own .env files
Capturas del proyecto
Imágenes reales detectadas en la documentación del repositorio.
Perfil de infraestructura
Estimación inicial derivada del repositorio y de los servicios detectados.
Despliegue detectado
Qué señales reales de instalación hemos encontrado.
To deploy pendpost locally: clone the repository and run `npm start` to build the dashboard on first startup. For production use, consider running a systemd service or Docker container with `docker compose up`. Always set `PENDPOST_MODE=production` in `.env` before deploying. Use an always-on host (e.g., a VPS) if you need publishing to continue when your computer is off. Never commit sensitive credentials to version control — use environment variables and the `.gitignore` file that pendpost provides.
Stack y dependencias
Componentes inferidos a partir de manifests, configuración y documentación.
Puede sustituir a
Relaciones detectadas por el análisis del repositorio y su categoría.
Experiencia real de la comunidad
Ayuda a contrastar la estimación automática con instalaciones reales.
Todavía no hay datos comunitarios para este proyecto.