banodoco/hivemind

A drop-in skill that lets a coding agent (Claude Code, Codex, etc.) search the Banodoco Discord message feed — a public, read-only PostgREST mirror of the Banodoco server. Banodoco is where a lot of the practitioners working on generative video & image tooling (Wan, Wan Animate, VACE, LTX, ComfyUI, Kijai's nodes, SCAIL, InfiniteTalk, training, etc.) actually talk to each other. This skill teaches your agent how to query it.

The endpoint exposes the same data the Discord shows — public messages, attributed by display name. Be respectful when surfacing tips and credit authors.

---

Install (paste this to your agent)

Install the hivemind skill: clone https://github.com/banodoco/hivemind and copy the hivemind/ folder into ~/.claude/skills/ (or your agent's skills directory). Then verify by running the smoke-test curl in SKILL.md. After that, use the skill whenever I ask about Banodoco / community best practices for any video or image generation tool.

That's it — the agent will do the rest.

Or run the one-liner yourself

git clone https://github.com/banodoco/hivemind /tmp/hivemind-skill-tmp \
  && mkdir -p ~/.claude/skills \
  && cp -r /tmp/hivemind-skill-tmp/hivemind ~/.claude/skills/ \
  && rm -rf /tmp/hivemind-skill-tmp \
  && echo "Installed. Restart Claude Code, then try: /hivemind"

For Codex or any agent that loads instruction files instead of skills, copy hivemind/SKILL.md into your AGENTS.md (or equivalent instruction file) — the content is self-contained.

---

What you get

• The endpoint URL + the public anon key (safe to commit — it's the publishable key, RLS makes it read-only).
• The table schema: content, author_name, channel_name, created_at, reactions, message_id.
• A taxonomy of which channels are high-signal (daily_summaries, wan_chatter, wan_comfyui, ltx_chatter, comfyui, *_resources) vs. background noise.
• A short list of power-users to weight (Kijai, Ablejones, djbfilmz, 42hub, BNDC the summary bot).
• PostgREST query patterns the agent can use directly: ilike substring search, in.(…) channel filtering, repeated-key AND, or=(…) for spelling variants like wan animate vs wananimate, author + time filtering.
• Gotchas — most importantly that fts (full-text search) times out on this table, so use ilike only.
• A "best practices for X" recipe: hit daily_summaries first, then Kijai-by-author, then topic channels, then cross-check wanx-troopers.github.io.

See hivemind/SKILL.md for the full contents.

---

Example prompts that will trigger the skill

• "What does Banodoco say about Wan Animate best practices?"
• "Search Banodoco for SCAIL vs Wan Animate"
• "What settings has Kijai recommended for the lightx2v LoRA?"
• "Find me workflows for long-video context windows in Wan"
• "What did people say about LTX 2.3 last week?"

---

Endpoint

GET https://ujlwuvkrxlvoswwkerdf.supabase.co/rest/v1/message_feed
Header: apikey: sb_publishable_O38oPBafrBoFrpi_rlWJvA_UJrulFsx

Smoke-test:

curl -s "https://ujlwuvkrxlvoswwkerdf.supabase.co/rest/v1/message_feed?select=content,author_name,channel_name,created_at&limit=3" \
  -H "apikey: sb_publishable_O38oPBafrBoFrpi_rlWJvA_UJrulFsx" | python3 -m json.tool

---

License

MIT. The data the endpoint serves is public Discord content authored by Banodoco members — credit them when you surface their tips.