Skip to main content

Watch And Change Workflow

Monitoring in Rug Radar is intentionally simple: store a baseline, take a new snapshot later, compute deltas, and trigger alerts if thresholds are crossed.

Why Use Monitoring​

Single scans answer "what is true right now?" Monitoring answers "what changed since I cared enough to save a baseline?"

This matters because many failures are not visible on the first check:

  • liquidity can disappear after entry interest arrives
  • holder concentration can worsen after redistribution
  • authorities can change
  • honeypot status can flip
  • risk can stack over several hours instead of all at once

Workflow​

1. Create a watch​

Prompt:

Watch <token address> with your thresholds

What happens:

  • Rug Radar normalizes thresholds
  • upserts a watchlist row in Postgres
  • runs analyze_token
  • stores the current analysis as the baseline snapshot

2. Re-check later​

Prompt:

What changed since last check for <token address>?

What happens:

  • Rug Radar loads the latest stored snapshot for the token
  • runs a fresh analyze_token
  • stores the new snapshot
  • computes deltas across selected metrics
  • evaluates watchlist thresholds
  • stores alerts that pass deduplication

3. Review the delta set​

The change report focuses on:

  • riskScore
  • liquidity
  • top10HoldersPercent
  • mintAuthority
  • freezeAuthority
  • honeypotStatus

One noisy check can mislead. Repeated deterioration is much more actionable.

Suggested Cadence​

  • High volatility: every 15-30 minutes
  • Normal conditions: every few hours

Default Thresholds​

If you do not supply custom thresholds, Rug Radar uses:

  • liquidityDropPercent: 20
  • riskScoreIncrease: 10
  • holderConcentrationIncrease: 5
  • alertOnAuthorityChange: true
  • alertOnHoneypot: true

Important Caveat​

This repository does not include a background scheduler. Watches are stored continuously, but alert evaluation only happens when get_token_changes runs.