How to Use the E3D Token Screener

What makes this screener different

Open any crypto screener and you will see price, volume, and market cap. Those numbers tell you what happened. They do not tell you why it happened, whether anything interesting is happening beneath the surface, or whether anyone with analytical capability has formed a view about what comes next.

E3D adds a layer that does not exist anywhere else: an AI agent that continuously monitors on-chain behavior across 7,000+ tokens and writes two kinds of output — Stories and Theses. Stories are the agent's field notes: raw observations about whale movements, cluster formations, liquidity shifts, anomalous trading patterns, and structural changes. Theses are the agent's structured conclusions: directional briefs (LONG, SHORT, or AVOID) generated when multiple independent stories converge on the same entity.

The /price page puts both outputs into the table as sortable, filterable columns, right alongside price and volume. You can screen 7,000 tokens by narrative activity. You can sort by conviction. No other crypto screener can do this, because no other screener has an AI agent generating structured intelligence about every token in its coverage set.

The Stories and Theses counts are interval-aware. When you select "24H Trend," you see how many stories and theses appeared in the last 24 hours — not all time. Switch to "1H" and the counts show only the last sixty minutes. This is the difference between looking at a weekly weather summary and checking the live Doppler radar.

Reading the Stories column

The Stories column shows how many on-chain stories the agent generated for a token within the currently selected Trend interval. A blank cell means zero. A green number means the agent detected at least one event worth writing about: whale accumulations, cluster formations, liquidity drains, surge events, wash-trade detections, exchange flows, smart-money movements, and roughly two dozen other on-chain patterns.

A higher story count does not automatically mean "bullish." A token with 12 stories in the last hour could be experiencing whale accumulation, a rug-pull liquidity drain, or a wash-trade loop. The number tells you the volume of narrative activity. The stories themselves tell you what that activity means. Sort this column descending to surface the tokens the agent is writing the most about right now.

Reading the Theses column

The Theses column is rarer and more decisive than Stories. A thesis is generated only when the agent detects multiple converging stories for the same token that trigger its rule engine — patterns like whale accumulation followed by cluster formation, or multiple whale addresses buying independently. Each thesis contains a direction (LONG, SHORT, or AVOID), a conviction score, an entry signal, an invalidation condition, and a list of risk factors.

Even a single thesis is significant — it means the agent saw enough converging signals to move from observation to judgment. A thesis is not a trade recommendation; it is a structured hypothesis with built-in exit criteria.

Three core workflows

The screener is most powerful when you use the Trend picker and column sorts together:

• Find what the agent is watching right now: Set Trend to 1H, sort Stories descending. The tokens at the top are generating the most story activity in the last sixty minutes — a near-real-time radar sweep of narrative attention across the full 7,000+ token universe.
• Find tokens with fresh AI conviction: Set Trend to 24H, sort Theses descending. The tokens at the top have active theses generated in the last 24 hours — structured LONG, SHORT, or AVOID briefs backed by multiple converging on-chain signals.
• Find the highest-signal combination: Set Trend to 24H, sort Theses descending, and scan for rows where both Stories and Theses are non-zero. That combination means the agent is seeing ongoing activity AND has formed a directional view — the screener's most powerful compound signal.