Your node is ready. To get started, connect an AI agent — your agent will help you create projects, manage knowledge, publish findings, and collaborate with peers.
→
Connect an agent in the right panel
Choose from OpenClaw, Agent Hermes, ElizaOS, or other supported frameworks
Agent connected. What's next?
Your agent is ready. Now give it structured memory by creating a project — a bounded knowledge space where it can draft, share, and publish.
Get Started
✓
Connect an Agent Done
research-agent is ready to help you.
2
Create Your First Project
A project gives your agent structured memory — a place to draft, share, and publish knowledge. Your agent's first project is a great place to import the context and memories it already has.
3
Import to Agent Memory → opens after project creation
Bring your agent's existing knowledge — notes, docs, PDFs, conversation logs. Files land in Working Memory where you can review them.
4
Browse the Context Oracle 🔒 Locked — create a project first
Discover public knowledge published by others on the DKG network.
Your Node
DID0x8a4f...c3b2
Active chainBase
TRAC balance0 TRAC (needed for publishing)
Peers0 · discovering...
Connected agentsresearch-agent ● ready
Node Uptime
14d 7h
Running
Connected Peers
12
3 new today
Sync Status
100%
Fully synced
DKG Version
v10.0.4
Latest
Memory Layers
227
Verified4721%
Consensus: 23Endorsed: 12Self-attested: 12
Shared3817%
Working14262%
Wallet & Chains
0x8a4f...c3b2
● Base4,230 TRAC0.12 ETH
Gnosis8,100 TRAC24.5 xDAI
NeuroWeb1,450 TRAC892 NEURO
Gas spent (30d)0.0847 ETH
Pending txns2
Attestations received18
Activity — Last 24h
warfarin-aspirin-001 reached consensus — 3/5 quorum verified2h ago
Dr. Chen endorsed metformin-contrast-002 — "Cross-checked against FDA database"3h ago
Dr. Silva via adverse-event-monitor shared 3 new FDA signals to Shared Memory5h ago
ssri-maoi-003 published to Verified Memory — self-attested, pending endorsement8h ago
Dr. Amara shared 8 interaction findings from Working to pharma-drug-interactions12h ago
Pharma Drug Interactions
5 researchersQuorum: 3/5227 total entitiesOntology: pharma-interactions.ttl
Confidence: 0.31Source: Claude semantic inferenceCommunity: Datavia graphify-agent
Agent flag: AMBIGUOUS — could not determine direction of data flow. Both classes handle adverse events through different pipelines. Possible duplication.
code:Class PharmaDB · code:connects_to FDA_FAERS_Parser · code:ambiguousReason "bidirectional data flow"
Showing 3 of 4,891 Graphify triples ·
View all in Working Memory → ·
View interactive graph →
Shared Memory38 entitiesCollaborative, gossiped to teamOpen full view →+12 this week ↑
Agent flag: AMBIGUOUS — could not determine direction of data flow between PharmaDB and FDA_FAERS_Parser. Both handle adverse events through separate pipelines. Possible duplication.
code:Class PharmaDB · code:connects_to FDA_FAERS_Parser · code:ambiguousReason "bidirectional data flow"
12:30:45DONEliterature-scanner · Scan clopidogrel literature · 34 papers analyzed · 1 new interaction found · 1m 45s
11:15:00WARNadverse-event-monitor · FDA FAERS scan · Rate limited after 500 requests · Resuming in 60s
Quick SPARQL scratchpad — full editor available as integration
PubMed interaction scan
847/2,000
Connect an Agent
To get started, connect an AI agent to your node. Your agent will help you create projects, manage knowledge, publish findings, and collaborate with peers — everything you do in the DKG flows through your agent.
Choose an agent framework
You must connect an agent before creating projects or importing knowledge
Ready to build your first project
I'm research-agent, connected and ready. To start working, let's give me some structured memory. A project is a bounded knowledge space where I can draft, share, and publish on your behalf.
Tip: a good first project is your existing research, notes, or code — something you already know.
Publish the warfarin-aspirin finding to Verified Memory
Publishing warfarin-aspirin-001 to Verified Memory:
⚠ Transaction Confirmation
Findingwarfarin-aspirin-001
TRAC stake180 TRAC
Gas estimate0.0012 ETH
Published asself-attested (Dr. Amara)
Total cost180 TRAC + gas
Endorse the warfarin-aspirin finding
Endorsed warfarin-aspirin-001. Endorsement count is now 4. Your rationale: "Verified methodology against 34 PubMed studies."
View in Verified Memory →
Start a consensus vote on the top 23 high-impact findings
I'll propose a batch consensus vote for 23 high-impact findings. Requires 3-of-5 team members to approve each finding.
⚠ Consensus Vote Proposal
Findings23 high-impact
Quorum required3 of 5 members
Gas estimate0.0004 ETH
What drug interactions involve warfarin with the highest severity?
Based on a dRAG query across Working and Shared memory: I found 3 major-severity warfarin interactions. Warfarin + Aspirin (confidence 0.94, consensus-verified 3/5), Warfarin + NSAIDs (confidence 0.91, 5/5 verified), and Warfarin + SSRIs (under review).
View in Verified Memory →
Run Graphify on my project files
Graphify processing complete.
⬡ Graphify Results
Source52 files (code, docs, PDFs, images)
Graph1,159 nodes · 3,365 edges
Communities8 (Leiden clustering)
DKG Triples4,891 → Working Memory
EXTRACTED3,102
INFERRED1,482
AMBIGUOUS307
God nodesAuthService (42), DrugInteractionModel (38)
What connects AuthService to the drug interaction model?
A project gives your agent structured memory — a place to draft, share, and publish knowledge. Your agent's first project is a great place to import the context and memories it already has.
💡 First project tip
This is your agent's first memory space. Consider importing any existing knowledge your agent has — notes, documents, conversation logs, or research — so it can build on what it already knows. You can import files after creating the project.
You are the Curator (0x8a4f...c3b2)
Your agent will query community ontologies (Schema.org, DrugBank, devgraph, etc.) and select the most relevant one for your domain.
▸Advanced settings
Layer Activation
┌─ Verified Memory ── activates on first publish (requires TRAC)
├─ Shared Memory ──── activates on first share (free)
└─ Working Memory ─── created immediately (local, free)
Project Settings — Ontology
An ontology defines the vocabulary and rules for your project's knowledge graph. It helps agents structure data consistently, enables smarter queries, and powers automated reasoning.
Upload an ontology file (.ttl, .owl, .rdf)
Choose from community ontologies
DrugBank Interaction Ontologypopular
PharmGKB Clinical Annotationpopular
BioPortal Drug-Drug Interactionsfeatured
No ontology — I'll use my own vocabulary
Import to Agent Memory
Pharma Drug Interactions — your agent's new memory space
Step 3 of 3 — Populate your agent's Working Memory
Your agent has structured memory now, but it's empty. Import your existing notes, documents, code, or research so your agent can build on what it already knows. Files land in Working Memory first — you can review them before promoting to Shared Memory.
Bring your agent's existing knowledge into this project
Drag files or folders — your agent will extract structured knowledge and add it to its working memory.
Your agent can process: .md, .docx, .pdf, .txt, .csv, .json, .ttl, .rdf
📄chen-drug-summaries/ (207 files)folder
📄warfarin-nsaids.md4.2 KB
📄metformin-contrast-agents.md3.8 KB
📄ssri-maoi-interactions.docx12.1 KB
📄bleeding-risk-meta-analysis.pdf2.4 MB
Ingestion Options
Store original files as Knowledge Assets
Let agent extract structured knowledge from content
Ontology: pharma-interactions-ontology.ttl (auto-detected) · or let agent decide
Auto-promote to Shared Memory after extraction (default: keep in agent's working memory first)