AI chats for running single-purpose private automations
NODES
Nodes are task-specific AI chat terminals
Each Node has one defined scope, one expected input pattern, and one workflow behind it.
Use Nodes when you want to run a focused automation instead of starting a general chat.
Nodes are used for:
- focused research
- crypto and market analysis
- contract or account review
- search-augmented answers
- reasoning and context compression
- prompt refinement
- workflow testing
- automation prototyping
A Node is not a general assistant. It is a constrained workflow interface.
This means that each Node accepts a defined input, runs a focused workflow, and returns a result inside the topic.
Nodes are used to test ideas, prototype workflows, run focused research, and collect feedback on useful automations.
WARNING
Nodes are currently under active development and may experience migration, downtime, or interface changes.
What Nodes Do
Nodes turn a user request into a structured workflow run.
A Node may:
- accept a specific input format
- call APIs or MCP tools
- run one Agent or Swarm
- perform a fixed sequence of steps
- return a summary, analysis, log, brief, score, or next-step output
- expose workflow details for review
Each Node topic explains what that Node does, what input it expects, what it can return, and what its limits are.
How to Use a Node
- Open a Node topic.
- Read what the Node does.
- Check the required input format.
- Start the chat.
- Send the requested input.
- Review the returned result.
- Continue only if the Node supports follow-up requests.
%%{init: {
"theme": "base",
"themeVariables": {
"background": "transparent",
"primaryColor": "#0f172a",
"primaryTextColor": "#ffffff",
"primaryBorderColor": "#38bdf8",
"lineColor": "#f59e0b",
"secondaryColor": "#111827",
"tertiaryColor": "#1e293b",
"fontFamily": "Inter, system-ui, sans-serif"
}
}}%%
flowchart LR
A["Open Node"] --> B["Read Scope"]
B --> C["Send Input"]
C --> D["Workflow Runs"]
D --> E["Agent / Swarm Uses Tools"]
E --> F["Result Returns in Topic"]
classDef user fill:#0f172a,color:#ffffff,stroke:#38bdf8,stroke-width:1px;
classDef run fill:#111827,color:#ffffff,stroke:#f59e0b,stroke-width:1px;
class A,B,C,F user;
class D,E run;
Why Nodes Differ from Standard Chats
| Standard chat | Node terminal |
|---|---|
| General conversation | Focused workflow execution |
| Broad instructions | Specific task rules |
| User manages the process | Workflow handles the process |
| One assistant tries to do everything | One scoped Agent or Swarm runs one job |
| Tools may be absent or inconsistent | Tools are selected for the Node’s purpose |
| Answer-focused | Output-focused |
| Usually returns a reply | May return briefs, logs, scores, summaries, or structured results |
Rules & Limits
- Only messages sent through the Node chat trigger workflow actions.
- Runtime, rate limits, model backend, memory, and tool access are set per Node.
- Some Nodes are free; some require member or premium access.
- Some Nodes use external APIs or MCP tools.
- Missing or invalid inputs may return errors.
- Outputs should be treated as workflow results, not final authority.
- Crypto, market, and contract outputs are research aids, not financial advice.
Quick Reference
Open Node → read scope → send required input → review result.
Navigate
Crypto
Use these Nodes for crypto research, market signals, contract review, and project tracking.
Inference
Use these Nodes for reasoning, context compression, inference chains, prompt refinement, and swarm-style analysis.
WebSearch
Use this Node for web research, source discovery, search-augmented answers, and external context gathering.