Data reads

Each read is a REST endpoint and, except asset positioning, an MCP tool with the same parameters and response. The API reference lists each read’s parameters, defaults, and fields.

Read	REST	MCP tool	Scope
Partner feed	`GET /v2/feed`	`list_feed`	`feed.read`
Trade events	`GET /v2/events`	`list_events`	`events.read`
Source messages	`GET /v2/messages`	`list_messages`	`messages.read`
Channel summaries	`GET /v2/channel-summaries`	`list_channel_summaries`	`summaries.read`
Aggregate summaries	`GET /v2/aggregate-summaries`	`list_aggregate_summaries`	`summaries.read`
Positions	`GET /v2/positions`	`list_positions`	`positions.read`
Open positions	`GET /v2/positions/open`	`list_open_positions`	`positions.read`
Trader directory	`GET /v2/traders`	`list_trader_directory`	`directory.read`
Trader activity	`GET /v2/traders/activity`	`list_trader_activity`	`stats.read`
Asset discovery	`GET /v2/assets`	`list_assets`	`directory.read`
Trader stats	`GET /v2/traders/stats`	`list_trader_stats`	`stats.read`
Asset stats	`GET /v2/assets/stats`	`list_asset_stats`	`stats.read`
Asset positioning	`GET /v2/assets/positioning`	None	`stats.read`
Trader rankings	`GET /v2/traders/rankings`	`rank_traders`	`stats.read`
Activity summaries	`GET /v2/activity-summaries`	`summarize_message_activity`	`stats.read`
Position Intelligence	`GET /v2/signals/*`	`get_*` tools	`stats.read`
Legacy trader discovery	`GET /v1/traders`	`list_traders`	`directory.read`

GET /v2/assets/positioning is pending release. Use it only after it is available in your environment.

Requests and responses covers the rules every read shares: identifiers, filters, pagination, time ranges, and ordering.

Which read answers which question

Question	Read
What is happening right now?	Partner feed
How are traders positioned on an asset now?	Open positions or asset positioning
Who is most active, or who performed best?	Trader rankings, with an explicit metric and time window
How many messages or events, and how is that trending?	Activity summaries
How did a trader perform?	Trader stats and positions
Which themes, narratives, drivers, or risks came up?	Generated summaries
Where did this event come from?	The event, then its `messageId` in messages

Generated summaries describe narratives. They are not evidence for counts, rankings, current skew, or consensus. For counts and rankings, use rankings and activity summaries, not pages of raw events. No single read answers top trader-asset pairs, price predictions, or “best trades” and “who to fade” without a defined sample and metric. An agent that combines reads for these should label its method, not present the result as an endpoint answer.

Combine and read results

Read	Know this
Trade events	Fetch an event’s position with `positionIds` on positions, and its source post with `messageId` as `ids` on messages. `messageId` can be null.
Source messages	`source.preview.text` is stored HTML. Sanitize it before you display it.
Trader activity	Rows come in alphabetical order of trader name, not by `positionCount`. `minPositionCount=0` includes traders with no positions, but an asset filter still requires a matching position.
Asset positioning	A holder counts once in `distinctHoldersCount`, even with several positions or both directions.
Activity summaries	Empty buckets are left out. `meta.totalMessages` and `meta.totalEvents` count every group, not only the returned ones.

Get started

Embed widgets

Build with the API

Receive webhooks

Connect AI agents

Which read answers which question

Combine and read results