Can AI replace posterly?

KINDA · weekend project

price $15/moyou'd save $180/yrbuild time multi-daycategory 🐦 social mediareplaced by 0 peoplebuild unverified

A basic compose-queue-publish loop against open-API networks (Mastodon, Bluesky, Telegram) is a weekend build, same as any scheduler. What's specific to posterly and harder to fake is the agent-native layer: an MCP server exposing tools like generate_captions, get_learned_voice, and find_available_slot so a coding agent can draft and post on your behalf, plus a voice model that's actually learned from your posting history and its own performance data rather than a one-off few-shot prompt. Add Google Business Profile review management, per-platform video transcoding, and client approval workflows for agencies, and the moat is upkeep and depth, not the initial compose/publish loop.

the prompt

Build me a personal social scheduler an AI agent can drive, to replace posterly. Requirements:

- Node + Express + better-sqlite3, with both a localhost compose page and
a small local HTTP API (list_slots, create_post, get_recent_posts) so a
coding agent can call it directly instead of only a human using the UI.
- Three open-API targets: Mastodon (access token), Bluesky (app password
via @atproto/api), and a Telegram channel (bot token). One adapter file
per network so a fourth is addable later.
- A /caption endpoint: given a topic, call an LLM API (key in .env) with
my last ~20 published posts as style examples, so drafts sound like me
instead of a generic AI voice. Cache the examples so it's not refetching
the DB on every call.
- Slot-based queue: posting times per weekday in my timezone, new posts
fill the next free slot, or pin an exact datetime; duplicate a post
across networks with per-network text.
- A node-cron tick every minute publishes due posts, marks each row sent
or failed before sending (no double-posts), retries failures 3 times.
- Attached images in media/ next to the database, resized per network
with sharp; a history page of the last 100 sent posts linking out.
- Runs on my always-on box; localhost only, no accounts, no telemetry.
- Out of scope: Instagram, LinkedIn, TikTok, Facebook, and Google Business
Profile (approval-gated APIs, say so in the README), review management,
client approvals, video transcoding, and analytics.
- README: getting a Mastodon token, a Bluesky app password, a Telegram
bot token, and pointing a coding agent at the local API.

$ open in your agent (prompt prefilled, you press enter) or copy it raw

why people still pay

Anyone can wire an LLM to a caption box; keeping a voice model current against real engagement data, holding pre-approved OAuth apps for the platforms that gate access behind business verification, and exposing a stable MCP/API surface an agent can drive reliably are all ongoing work, not a one-time build. Agencies additionally pay for the client-review layer, which needs its own auth and approval state machine.

what you lose

x18 maintained platform integrations, including approval-gated ones (Instagram, LinkedIn, TikTok, Facebook, Google Business Profile)

xa voice model trained on your actual post history and what performed well, not a static prompt template

xGoogle Business Profile review replies and photo management

xclient approval portal and content-plan review for agencies managing multiple brands

xper-platform video/image transcoding tuned to each network's limits

xa hosted MCP server + REST API built for agents, not bolted on after the fact

prior art · use these instead of building, if you'd ratherPostizOpen-source social scheduler, self-hostable, covers much of the DIY core for open-API networks.↗