SkillFoxx

Hivemind skill

A skill that searches the Banodoco community corpus for video and image generation tips on Wan, LTX and ComfyUI.

  • 76 ★ on the repository
  • Last commit
  • MIT
Skill

Extended access

Why this level

  • Queries go to a third-party public community server.
  • The write path accepts a contributor key and changes the shared corpus.
All reasons and checks

banodoco/hivemindAuthor:banodoco

Install

In your terminal, with SkillFoxx CLI

npx skillfoxx add skills/banodoco-hivemind

Detects the agents on your machine, shows the access level and pins the version.

Other ways to install

Access and checks

How we review

Extended accessThe tool runs code or makes network calls and only reads external data.

  • Queries go to a third-party public community server.
  • The write path accepts a contributor key and changes the shared corpus.

This is third-party code. Review the repository files before installing.

What it does

Hivemind gives a coding agent the shared knowledge corpus of the Banodoco community: the Discord message feed, external resources such as articles, transcripts and workflows, and curated answers with sources. The agent finds practice missing from model cards and READMEs: settings, comparisons, pitfalls and links to member builds. Search runs through the `python3 executors/search/run.py` script, which scans three scopes in parallel and ranks the result on the client. Filters narrow a search by channel, author or resource kind. Writing back goes through a separate path, the contribute function, which needs a contributor key.

Who it is for. People working with generative video and images who want live community hints instead of official documentation.

Good fit when

  • You pick settings for Wan Animate, VACE or LTX and want others experience.
  • You look for a ComfyUI workflow from well-known node authors.
  • You want to know what the community thinks of a model before installing it.
  • A researched answer should be saved to the corpus for future queries.

Not a fit when

  • You need official model parameters, better taken from the author documentation.
  • There is no network access, since the corpus lives on a remote server.
  • The question is not about video or image generation, the only area the corpus covers.

Example request

What does the Banodoco community advise for Wan Animate on long clips? Find messages and links to workflows.

Limitations

The skill reads a public server on the Banodoco side: queries are visible to the corpus owner and availability depends on the author. The content is collected from member messages, so some tips are outdated or disputed. Evidence in the corpus is marked as reported or observed. Text search over the unified feed times out, so the skill recommends searching individual tables. Writing to the corpus requires a contributor key, and editorial decisions belong to editors only.

How to disable. Delete the skill folder from the agent skills directory. If you were issued a contributor key, store it separately and remove it from the environment.

FAQ

How do I search one channel?

Add the `--channel` flag, for example `python3 executors/search/run.py --query "lora" --channel wan_chatter`. The same search can be limited by author.

Where does the content come from?

From the public Banodoco corpus: Discord messages, external resources such as articles and video transcripts, and curated answers with cited sources.

How does writing to the corpus work?

Through `POST /functions/v1/contribute` with the `X-Contributor-Key` header. Submitting a resource, proposing a revision and recording evidence are available.

Is every edit accepted?

No. A new resource revision waits for an editor decision, only the accepted one ranks, and rejected or historical revisions never appear in search.

Are external packages required?

The search executors use only the Python standard library, so no packages are needed for reading.

Discussion

No questions or reviews yet.

Skills

A skill that writes prompts for video, image and music generation on 14 platforms and can submit them in the browser

SkillBroad access274

A Codex skill for images and video through the paid Higgsfield API with cost estimates, a spend cap and result downloads.

SkillExtended access68

Clipmivo Video

clipmivo-video

Official

A skill for ClipmivoAI: the agent quotes a video task in credits, starts generation, follows it and downloads the result.

SkillExtended access141

A skill for knowledge explainers: the agent researches a topic, writes screen copy and builds a HyperFrames project for MP4

SkillExtended access61

watch-cli is a skill and CLI that hands an agent the frames and transcript of any social video so it can read the clip itself.

SkillExtended access260

A skill and CLI that let Claude Code or Codex make videos as web pages, with the Studio editor to change the result

SkillExtended access6.4K

README in short

The repository describes a shared corpus with a public read path and a locked write path. Knowledge is stored as resources with immutable revisions, and search works on the accepted revision. Three ingestors publish articles, ComfyUI workflows and YouTube transcripts in one format.

README badge

Original SKILL.md

---
name: hivemind
description: >
  Search the Banodoco knowledge corpus — a public PostgREST endpoint
  combining a Discord message feed, external resources, and curated
  distillations. Covers generative video/image tooling (Wan, Wan Animate,
  VACE, LTX, Comfy, Kijai's nodes, SCAIL, InfiniteTalk, training, etc.).
  Use this whenever the user asks "what does Banodoco say about X",
  "best practices for a model", "what are people doing with a tool",
  "what settings did a named person recommend", or wants real-world tips that
  aren't on the model card / README. Channels like daily_summaries,
  wan_chatter, wan_comfyui, ltx_chatter, comfyui, *_resources are the
  goldmine. For contributing back, use the write path via the contribute
  edge function.
---

# hivemind

A read-only public endpoint exposing the Banodoco knowledge corpus —
community knowledge about video/image generation you can't get from official
docs: workflow tips, model comparisons, settings tweaks, gotchas, and links
to Kijai/Ablejones/community workflows. It layers three things into one
searchable feed: **messages** (raw Discord), **resources** (articles,
transcripts, workflows), and **distillations** (curated Q&A with cited
sources). Distillations make the corpus self-improving — every researched
answer you submit becomes permanently searchable.

## Quick start

```
Endpoint: https://ujlwuvkrxlvoswwkerdf.supabase.co/rest/v1
Key (anon publishable, safe to commit):
  apikey: sb_publishable_O38oPBafrBoFrpi_rlWJvA_UJrulFsx
```

Smoke test (raw search surface — per-token OR on message content; do NOT
text-search `unified_feed`, the UNION scan times out with 57014):

```bash
curl -s "https://ujlwuvkrxlvoswwkerdf.supabase.co/rest/v1/message_feed?select=message_id,content,author_name,channel_name,created_at&limit=5&order=created_at.desc&or=(content.ilike.*wan*,content.ilike.*animate*)" \
  -H "apikey: sb_publishable_O38oPBafrBoFrpi_rlWJvA_UJrulFsx"
```

### Surfaces at a glance — pick by what you want

| You want to… | Use | Why |
|---|---|---|
| Search curated Q&A (distillations) | `distillations` + `or=(question.ilike.*T*,answer.ilike.*T*,conditions.ilike.*T*)` | flywheel, highest signal; status `pending,approved` |
| Search resources (workflows, articles) | `external_resources?kind=eq.<resource-kind>&or=(title.ilike.*T*)` | raw table, kind btree + trigram GIN |
| **Search message content** (free text) | `message_feed` + `or=(content.ilike.*T1*,content.ilike.*T2*)` | per-token OR is index-friendly; `unified_feed` ilike **times out** (57014) |

Report a problem