How filters work
Spark checks every new token against your active filter set. A token only reaches the feed if it clears every active condition.
Core rules
A disabled filter group has no effect. Only turn on what your strategy actually needs.
The filter groups
| Group | Asks |
|---|---|
| Dev Holdings | How much of supply is the developer holding right after launch? |
| Migration | How often do this dev's tokens graduate to a real DEX? |
| Funding | How was the launch wallet funded - fresh, old, mixed? |
| Protocols | Which launchpad - pump.fun, bonk.fun, or both? |
| Fees | Total / average / per-mode fees on the dev's previous tokens. |
| Community | X (Twitter) community attached? Members? Creator followers? |
For a starting point that works without manual tuning, see Filter presets - three ready strategies you can import in one click.
How to apply a preset
- Settings → Main - set values manually, or
- Settings → Config → Import from File - paste a JSON preset.
- Watch the feed for an hour. Too empty → loosen. Too noisy → tighten.
Per-token feedback
Spark surfaces every active filter result on the token card itself, so you can see at a glance why a token passed (or, when running in debug mode, why it failed). This makes filter tuning a feedback loop, not guesswork.