docs: social platform integration feasibility research

[amantu-qbit]

What & why

Research deliverable for roadmap items #12 "Post Builder and Publish — YouTube" and #13 "LinkedIn, X, Telegram, Viber, YouTube integration". Adds docs/social-platform-integration-research.md — an in-depth feasibility study of whether these features are possible and what can realistically be built, given that Palworld Server Manager is a distributed desktop app with no always-on backend.

No product code changes — documentation only.

Key findings

• The governing constraint is architecture, not any one API. A distributed open-source desktop app can't safely ship a client secret, isn't always running, and keeps credentials local. That decides feasibility before any single platform does. Recommended model: bring-your-own-credentials (no shipped secret), the always-on Bridge as host for event-driven/scheduled posting, and no mandatory backend.
• Three distinct capabilities are bundled under "integration" and shouldn't be treated as one: broadcast/publishing, event-driven notifications, and two-way chat relay.
• Per-platform verdicts:
  • 🟢 Telegram — free Bot API, bring-your-own token, no OAuth/review. Ship first.
  • 🟢 Discord (not in the original list) — webhooks need no OAuth; likely the highest-value integration for a gaming community. Recommend adding to scope.
  • 🟡 X / Twitter — feasible via OAuth2 PKCE, but pay-per-use is the default for new developers since Feb 2026 (~$0.015/post, ~$0.20 with a link); cost- and churn-aware.
  • 🟡 YouTube video upload — feasible via PKCE + BYO Google project; upload quota cost dropped ~16× in Dec 2025. YouTube community posts have no supported API.
  • 🔴 LinkedIn — requires a confidential client (backend) + a registered legal entity + app review. Not practical for this app.
  • 🔴 Viber — bots are commercial-only since Feb 2024. Not viable for the audience.
• Item #12 "Post Builder" is scoped as a compose-once/fan-out UI with per-platform previews, templates/variables, map-screenshot export, and scheduling delegated to the Bridge.
• A phased plan (zero-OAuth Telegram/Discord + notifications first, opt-in X/YouTube broadcast second, hard cases reassessed last) closes the doc.

Findings are compiled against then-current (2026-07) platform terms with sources cited inline; pricing/policy on these platforms changes often and should be re-verified before implementation.

🤖 Generated with Claude Code

https://claude.ai/code/session_01HWMnvsLFBFWpZwHNjjnk69

---

Generated by Claude Code