HashPilot’s AI features are powered by Anthropic’s Claude. Every AI endpoint reads a single environment variable —Documentation Index
Fetch the complete documentation index at: https://mintlify.com/natureloved/HashPilot/llms.txt
Use this file to discover all available pages before exploring further.
ANTHROPIC_API_KEY — so setup is straightforward whether you’re running locally or deploying to Vercel.
Get your API key
Create an Anthropic account
Go to console.anthropic.com and sign up or log in.
Create a new API key
In the Anthropic Console, navigate to API Keys and click Create Key. Give it a descriptive name (for example,
hashpilot-local).Copy the key
Copy the key immediately after creation. Anthropic will not show it again. If you lose it, you will need to create a new one.API keys follow the format
sk-ant-api03-....Setting the key in Vercel
For production deployments on Vercel, environment variables are set through the project dashboard — not in code.Open your project settings
In the Vercel dashboard, go to your HashPilot project and click Settings → Environment Variables.
Add the variable
- Name:
ANTHROPIC_API_KEY - Value: your API key (
sk-ant-api03-...) - Environment: select Production (and optionally Preview)
Which endpoints use the key
All four AI endpoints readANTHROPIC_API_KEY from process.env. None of the endpoints cache or store responses.
/api/chat — AI Strategy Chat and Archetype roadmap
/api/chat — AI Strategy Chat and Archetype roadmap
Powers the real-time streaming chat interface in AI Strategy Chat and the Miner Archetypes AI roadmap generator. Uses
claude-sonnet-4-6 with a max of 1500 tokens per response. Responses stream token-by-token to the browser.If the key is missing or empty, this endpoint returns HTTP 500 with "API configuration missing."./api/oracle — Oracle scenario readings
/api/oracle — Oracle scenario readings
Powers the Claim Oracle decision engine. Receives a scenario matrix (network growth, price move, new players, halvings) and returns a structured verdict: BULLISH / BEARISH / UNCERTAIN, plus a recommended action. Uses
claude-sonnet-4-6 with a max of 1000 tokens per reading./api/fomo-story — FOMO Machine narrative debrief
/api/fomo-story — FOMO Machine narrative debrief
Generates the narrative debrief in the FOMO Machine after the counterfactual simulation runs. Describes opportunity cost in the terminal’s retro-tactical voice. Uses
claude-sonnet-4-6./api/digest — Daily intel digest
/api/digest — Daily intel digest
Populates the Daily Digest card on the Dashboard with a fresh tactical summary. Uses
claude-sonnet-4-6 with a max of 1000 tokens per digest.Usage and cost
All endpoints use theclaude-sonnet-4-6 model. Per-request token limits are:
| Endpoint | Max tokens |
|---|---|
/api/chat | 1500 |
/api/oracle | 1000 |
/api/fomo-story | 1500 |
/api/digest | 500 |
Next steps
Run locally
Set up and start the dev server on your machine.
Deploy to production
Push to Vercel and configure the key in your project settings.