title: cluas_huginn
emoji: 💬
colorFrom: yellow
colorTo: purple
sdk: docker
pinned: false
hf_oauth: true
hf_oauth_scopes:
- inference-api
license: apache-2.0
short_description: A gathering of guides, a council of counsels
An example chatbot using Gradio, huggingface_hub, and the Hugging Face Inference API.
the above is just a template readme they gave me when I created this space on hugging face, feel free to ignore it.
🐦⬛ Cluas
A gathering of guides, a council of counsels
Clúas (Gaelic: "ear") — a dialectic research tool where AI experts collaborate, remember, and build knowledge over time.
What it Does
Corvid Council is a multi-agent MCP system that enables dialectic research through collaborative AI agents. Present a question to the council and watch as four specialized experts research, debate, and synthesize knowledge—building on past discussions through shared memory.
Modes
- Collaborative Mode: Ask a question and receive synthesized research with sources
- Active Mode: Join the discussion, steer research, challenge claims
Dialectic Process
- Thesis: Characters present initial findings using specialized tools
- Antithesis: Characters challenge, verify, and provide counter-evidence
- Synthesis: Council builds consensus and adds to collective memory
- Evolution: Future discussions build on accumulated knowledge
Why It Matters
Most AI assistants are stateless. cluas_huginn Council remembers, learns, and builds knowledge over time.
Characters
- Corvus: 🐦⬛ (black bird)
- Magpie: 🪶 (feather, playful)
- Raven: 🦅 (eagle, serious)
- Crow: 🕊️ (dove, peaceful observer)
Taglines
- "A gathering of guides, a council of counsels"
- "Research that remembers, knowledge that accumulates"
- "Multi-agent MCP research collective"
v2:
Cluas Huginn: Multi-Agent Dialectic System
What It Does
Four AI agents with distinct roles debate questions using structured dialectic.
Architecture
- Corvus: Academic verifier (literature search)
- Raven: Accountability enforcer (news verification)
- Magpie: Trend explorer (pattern connector)
- Crow: Grounded observer (environmental data)
Key Innovations
- Unified inheritance architecture
- Shared epistemic principles with character differentiation
- Tool-use heuristics per character
- Steelmanning and collaborative disagreement built-in
Tech Stack
- Base: Python, Gradio
- LLMs: Groq/Nebius (Qwen 3)
- Tools: Academic search, news verification, web exploration
- Memory: Persistent character memories
What Makes This Different
- Not just multiple LLMs - distinct epistemic roles
- Structured dialectic (thesis/antithesis/synthesis)
- Tool usage guided by character personality
- Collaborative, not adversarial
Hackathon Track Entries
"building-mcp-track-enterprise" → Track 1
"building-mcp-track-consumer" → Track 1
"building-mcp-track-creative" → Track 1
"mcp-in-action-track-enterprise" → Track 2
"mcp-in-action-track-consumer" → Track 2
"mcp-in-action-track-creative" → Track 2
This project is licensed under the Apache 2.0 License.