NakamotoNotes Bitcoin Barometer
# NakamotoNotes Bitcoin Barometer — MCP server
Read-only MCP server for the NakamotoNotes Bitcoin Barometer: Heat Score 0-100, zone, BTC price with 24h/7d change, the ten indicators with value and heat, and daily history.
## Install
- **Hosted (no install):** add `https://nakamotonotes.com/mcp` as a remote MCP server (Streamable HTTP, no auth).
- Claude: one click from the [Claude directory](https://claude.ai/directory/nakamoto-notes-bitcoin-barometer).
- Grok: grok.com/connectors > New Connector > Custom, paste the URL.
- Cursor and Grok Bot: install the `nakamotonotes-bitcoin-barometer` plugin (manifest in `.cursor-plugin/plugin.json`, server in `mcp.json`).
- **Claude Desktop (one click):** download `nakamotonotes-barometer.mcpb` from [nakamotonotes.com](https://nakamotonotes.com/downloads/nakamotonotes-barometer.mcpb) and open it.
- **Any MCP client (stdio):** `npx -y @nakamotonotes/mcp`
```json
{ "mcpServers": { "nakamotonotes": { "command": "npx", "args": ["-y", "@nakamotonotes/mcp"] } } }
```
Tools: `get_market_snapshot`, `explain_indicator`, `get_history`. No API key, no personal data; it reads the public feed at https://nakamotonotes.com/open/barometer.json.
Data licence: CC BY 4.0, attribute NakamotoNotes. The reading describes the current state against its history; it is not a forecast or investment advice.
TDQS
Scored across 3 tools
Each tool has a distinct angle: get_market_snapshot is the current full reading, get_history is a time series, and explain_indicator is a deep dive on a single indicator. There is mild overlap since the snapshot already includes indicator values and one-line meanings, but the descriptions distinguish 'current overview' from 'explanation' clearly enough.
All names are snake_case verb_noun, but two use the get_ prefix while explain_indicator uses explain_. The deviation is minor and readable, so the pattern is still mostly predictable.
Three tools is on the lean side, but for a read-only, descriptive barometer the small surface is well-scoped: one for the current reading, one for history, one for indicator education. Nothing feels redundant.
The lifecycle of a read-only data source is largely covered: current snapshot, historical series, and per-indicator explanation. Minor gaps remain, such as listing all valid indicator names or fetching a full snapshot for a past date, but agents can work around these via get_history.