Sentimentor AI research agent · Augur

Sentimentor is an AI market-research terminal—not only a live-data monitor.

The terminal combines multi-asset charts, screeners, news, sentiment, institutional activity, options, fundamentals and public-record research with an integrated agent called Augur. Augur can turn a finding into editable PineNode code you can test, validate and—when eligible—forward-test through Sentimentor simulation or a connected Alpaca Paper account.

Forward research ledger

Test whether the answer predicts anything—not only whether a backtest looks good.

Research Evals preregister Augur's prompt, model lineage, universe, evidence cutoff, ranked picks and probabilities. Server-captured entry prices are sealed with the record, then a scheduled scorer measures return and calibration against the chosen benchmark after the full trading horizon.

01

Blind forward selection

02

Historical replay

03

Regime testing

04

Human/tool comparison

05

AI-vs-rules ablation

Evaluation methodology

The ledger measures research productivity and forward performance separately. Historical evidence manifests are declarations unless independently verified; no result is a guarantee or financial advice.

Workflow distribution

Inspect, subscribe, or fork.

Sentimentor exposes three distinct actions instead of treating them as the same thing. Built-in preset automations have scheduled delivery; community workflows have editable ownership and provenance.

Inspect

Review a preset or community workflow’s pipeline, research tools and output before using it.

Subscribe / Unsubscribe

Receive or stop a built-in preset’s server-scheduled email result. The preset remains centrally maintained.

Add to my workflows

Import a community workflow as an editable private fork, then run, schedule or republish your variation.

See all eight preset automations

The research loop

One conversation, from question to evidence.

Augur is not a general chatbot guessing from model memory. The agent builds a research plan, dynamically selects tools, evaluates their results over multiple rounds and preserves the path from current evidence to executable logic.

01

Ask

Describe a ticker, market question, screen or trade hypothesis in plain English.

02

Research

Augur can decompose the investigation into an auditable workflow of data, transform, scoring and branch operations.

03

Generate

Each operation stays inspectable, and the terminal step emits readable, editable PineNode strategy code.

04

Verify

Screen on discovery data, test across rolling validation folds, open one final lockbox, disclose recorded trials and Monte Carlo downside, then share the result.

05

Paper execute

Collect exploratory or validated forward evidence with simulated fills, or optionally use a connected Alpaca Paper balance for eligible long-only U.S.-equity runners.

Open-ended analysis, decomposed instead of hidden.

For complex quantitative requests, Sentimentor can represent the investigation as a directed workflow with custom data steps, filters, scores and conditional branches. Every tool-backed operation exposes its code-shaped call and result. A required terminal operation then generates the complete PineNode strategy, preserving a traceable path from research inputs to code that can be backtested and verified.

Research coverage

The 41 tools are callable research capabilities, not a count of upstream feeds. A tool count and a data-source count measure different things. Augur uses these capabilities to investigate the research categories below; availability and latency can vary by plan and source.

Price, charts & screeners

Live snapshots, historical bars, movers, market overviews and multi-factor equity screens.

OHLCV · screeners · market breadth

Options & institutional flow

Options flow, IV rank, put/call trends, dark-pool levels and volume delta.

sweeps · IV · dark pools · volume delta

Fundamentals & positioning

Financial metrics, analyst recommendations, executive compensation, institutional ownership and insider activity.

fundamentals · executive pay · 13F · Form 4

Public, regulatory & credit evidence

Congressional trading, lobbying, contracts, filings, patents, solvency, debt maturities and credit-facility evidence.

Congress · Altman Z · debt maturities · credit evidence

News, sentiment & social

Ticker news, Sentimentor scores, directional predictions, Reddit buzz and X trends.

news · sentiment · Reddit · X

Macro & corporate events

Economic, earnings, IPO and ex-dividend calendars, CFTC positioning and corporate activity.

FOMC · earnings · CFTC · corporate actions

A concrete example

Research that ends in something testable.

Ask Augur

“What were the most-traded stocks by Congress last month? Analyze the top candidate and create a trend-confirmation strategy.”

  • Queries normalized congressional and public-record data.
  • Ranks the results and enriches the leading ticker with market, sentiment and flow context.
  • Shows the evidence as typed tables and cards instead of opaque prose.
  • Generates PineNode source that opens directly in the editor for review and testing.
strategy.tsEditable
const fast = new EMA(strategy.param('Fast EMA', 50, 20, 100, 10));
const slow = new SMA(strategy.param('Slow SMA', 200, 100, 300, 20));
const atr = new ATR(strategy.param('ATR Period', 14, 7, 30, 7));

strategy
  .name('MSFT Congressional Candidate Trend Confirmation')
  .onBar(({ series, buy, sell }) => {
    fast.next(series);
    slow.next(series);
    atr.next(series);
    if (series.length < 210) return;
    const close = series.close(0);
    const priorClose = series.close(1);
    const volume = series.volume(0);
    const avgVolume = Array.from({ length: 20 }, (_, i) => series.volume(i + 1)).reduce((a, b) => a + b, 0) / 20;
    const breakout = Math.max(...Array.from({ length: 20 }, (_, i) => series.high(i + 1)));
    if (close > slow.get(0) && fast.get(0) > slow.get(0) && close > breakout && volume > avgVolume * 1.2) buy('Uptrend breakout with volume confirmation');
    if (close < slow.get(0) && priorClose >= slow.get(1)) sell('Confirmed close below 200-day trend');
  });
Backtest
Validation + final lockbox
Share /verify
Alpaca Paper

An answer is a starting point. Verification and forward evidence make it useful.

Generated code remains fully visible and editable. AI-assisted code remains exploratory until its exact configuration completes discovery, rolling validation, a final lockbox, Monte Carlo and the server-recorded trial protocol. Broker-integrated with Alpaca Paper today. Live-money trading is planned for December 2026, with additional brokers to follow. Sentimentor is not a broker. Alpaca Paper uses simulated funds and fills that can differ from live-market execution. Backtests and paper results are evidence—not forecasts or financial advice.

Compare Sentimentor with AI-powered stock analysis platforms