> ## Documentation Index
> Fetch the complete documentation index at: https://api.unusualwhales.com/docs/llms.txt
> Use this file to discover all available pages before exploring further.

> ## Agent Instructions
> API requests use the base URL https://api.unusualwhales.com and require a bearer token in the `Authorization` header (`Authorization: Bearer <API_KEY>`). Create and manage API tokens at https://unusualwhales.com/dashboard/api.
> For live market data inside an AI tool, use the Unusual Whales MCP server at https://unusualwhales.com/public-api/mcp.
> Instructions for agents using Unusual Whales tools: https://unusualwhales.com/skill.md

# predictions-insiders

> Prediction market insider signals based on signal such as new account withs low activity, taking unusually large positions and more.

Topic
messages**1**
routing**single type**
encoding**protobuf**

## Messages on this topic

[all messages
PredictionsInsider
A potential insider signal: a fresh, low-activity account taking an unusually large position in a single market. Published to the `predictions-insiders` Kafka topic each time insider detection runs; an existing signal is re-published when its underlying position changes. A position is flagged when all of the following hold: - its cost basis is at least $1,000, at an average entry price of at most $0.90 per share, and its realized loss is no worse than -\$100 - the account holds fewer than 5 distinct positions - the account's first observed trade was fewer than 20 days ago
17 fields→](/docs/kafka/types/PredictionsInsider)

All messages on this topic are of a single type.

## Consume the topic

```python Python theme={null}
from confluent_kafka import Consumer
from unusualwhales.schema_pb2 import PredictionsInsider

# CONFIG comes from the Connecting section of the Kafka overview.
consumer = Consumer(CONFIG | {"group.id": "my-predictions-insiders-consumer"})
consumer.subscribe(["predictions-insiders"])

while True:
    msg = consumer.poll(1.0)
    if msg is None or msg.error():
        continue

    event = PredictionsInsider()
    event.ParseFromString(msg.value())
    print(event)
```