Skip to content

Social Radar SOP

Purpose

The social radar surfaces non-watchlist tickers with notable Discord social signal activity. The CLI provides filtered data; the agent decides what to do with it.

Data Source

Radar data appears in the weekly review artifacts as a table:

| Ticker | Signals | Channels | BUY | SELL | Strong | Sector | Top Channel (WR) |

It can also be fetched on demand: 

openfin social radar --json

Reasoning Steps

1. Assess Signal Quality

For each radar ticker, evaluate: - Channel count: 2 channels is the minimum filter. 4+ with high WR is notable. - Win rate: At least one channel with 50%+ WR passed the filter. Look for 80%+ WR channels. - Signal strength: High STRONG count relative to total signals indicates conviction. - Directional clarity: Is the BUY/SELL ratio clearly directional, or mixed?

2. Match to Existing Theses

Compare radar tickers against active thesis narratives: - Does the ticker operate in a sector covered by an existing thesis? - Does it compete with, supply to, or benefit from the same macro trend? - Would adding it strengthen an existing thesis as a secondary or hedge symbol?

If a match exists: 

# Add to existing thesis
openfin thesis save --slug <existing-slug> --add-symbol <TICKER>:secondary:starter

3. Identify Clusters

Look for multiple radar tickers in the same sector or theme: - 3+ crypto tickers → potential crypto thesis - Multiple EV/battery tickers → potential EV supply chain thesis - Healthcare cluster → potential biotech/pharma thesis

If a cluster warrants tracking: 

openfin thesis save --slug <new-slug> --status draft --title "..." --narrative "..." \
  --symbols TICKER1:primary:starter,TICKER2:secondary:starter

4. Quick-Track Individual Tickers

For a single high-signal ticker that doesn't fit a thesis: 

openfin watchlist add <TICKER>

This adds it to the adhoc watchlist for data collection. It will appear in future scoring without needing a full thesis.

5. Dismiss Noise

Tickers with high signal count but no high-WR channels, or signals concentrated in a single low-quality channel, can be dismissed. No action needed — just note the reasoning.

Important

  • Draft theses are excluded from the watchlist and scoring pipeline. They exist as proposals for human review.
  • The agent recommends but does not promote drafts to active. Promotion is a human decision.
  • Radar is data, not a recommendation. The agent reasons about it; the CLI just surfaces it.