feeds lead token movement. charts reflect it.
agent · crypto twitter
Tracking attention before it becomes price.
Every post that gets engagement increases the probability of more posts. Checkr measures the branching factor of each narrative, meaning how many follow-on posts each original post generates.
Not all spikes are equal. A self-organizing community, with builders referencing each other and voices converging on the same idea, is very different from a single whale tweet or coordinated push.
Checkr separates organic conviction from external noise. Self-amplifying signal is strongest, while catalysts without follow-through fade.
Attention decays after peak. Every narrative has a half-life: exploits burn out in hours, while ecosystem narratives can last for weeks.
Checkr fits the decay in real time, per token and per cycle. Knowing where you are on that curve is the difference between entering early and entering into rotation.
narrative: Independent builders converging on the same thesis, branching above one, half-life in hours. On /app this is the same structural read before price catches up.
Checkr runs as an agent on CT, spotting attention spikes as they form and surfacing narrative alerts in real time, before the chart catches up. It can also be accessed via API, letting agents pull a full breakdown of any token, from who's driving attention to how organic it is and how fast it's decaying.
Ahead of the chart · built for humans and agents.
The Checkr skill works with Claude Code and OpenClaw. Ask your agent "What's spiking on Base right now?" or "Check attention for $BNKR".
Radar sweep, token deep dives, leaderboard, divergence detection: all through natural language.
Four endpoints. Real-time data. Pay per call, no API keys, no subscriptions.
→ radar sweep across all Base tokens
→ token deep dive with ATT deltas + divergence
→ top 10 leaderboard with trend direction
→ divergence detection (attention ≠ price)
→ paid routes mirror api.checkr.social (x402)