The InDecision Signal API: Programmatic Access to the 6-Factor Framework

The InDecision Framework has been running live analysis since 2019. Every signal it produces — the directional bias, the conviction score, the per-factor breakdown — has lived inside a Python engine on a Mac Mini, consumed by a Discord bot and a Polymarket trading system. The output was always structured data. It was never accessible as structured data.

That changes with the Signal API.

What the API Actually Returns

The Signal API exposes a single endpoint that returns the current InDecision analysis for a given asset. No authentication dance, no SDK dependency, no WebSocket subscription model. A GET request returns a JSON object.

{
  "asset": "BTC/USD",
  "timestamp": "2026-02-26T14:30:00Z",
  "bias": "BEARISH",
  "conviction_pct": 74.3,
  "factors": {
    "daily_pattern":      { "score": 22.5, "max": 30, "signal": "BEARISH" },
    "volume":             { "score": 19.8, "max": 25, "signal": "BEARISH" },
    "timeframe_alignment":{ "score": 14.0, "max": 20, "signal": "BEARISH" },
    "technical":          { "score": 10.5, "max": 15, "signal": "NEUTRAL" },
    "market_timing":      { "score": 7.5,  "max": 10, "signal": "BEARISH" }
  },
  "bull_case": 25.7,
  "bear_case": 74.3,
  "spread": 48.6,
  "risk_context": {
    "gate": false,
    "flags": []
  },
  "meta": {
    "engine_version": "2.1.0",
    "exchange": "coinbase",
    "timeframe": "1d"
  }
}

Every field maps directly to the scoring engine. conviction_pct is the winning case score. spread is the gap between bull and bear cases — the confidence margin. Each factor returns its raw score against its maximum possible weight, plus the directional signal that factor produced independently.

The risk_context object acts as a gate. When gate is true, the conviction score is capped regardless of factor alignment — meaning a high-impact macro event (FOMC, CPI, major protocol upgrade) has been detected and the framework is signaling caution.

Who This Is For

Discord community operators. If you run a crypto trading community, you've built your signal delivery on screenshots, manual write-ups, or expensive third-party data feeds. The Signal API delivers structured conviction data directly into your bot. Format it however your community expects. Update it on whatever cadence makes sense. The data is machine-readable — the presentation layer is yours.

Dashboard builders. The conviction score, factor breakdown, and bull/bear spread are designed for visual representation. A gauge, a radar chart, a traffic-light grid — the API returns the components, not the visualization. Build the interface your users need.

Automated alert systems. The most valuable signal InDecision produces is often the transition — when conviction crosses from NEUTRAL to BEARISH at 70%+, or when a previously aligned bias flips. Polling the API on a schedule and triggering alerts on state changes is a straightforward integration.

Trading tool developers. If you're building systematic trading infrastructure, the Signal API provides a conviction layer that sits between raw market data and execution logic. Your system handles entries, exits, and position sizing. InDecision handles the directional read and confidence assessment.

What the Factor Breakdown Enables

Most signal services deliver a single number or a direction. Buy. Sell. 7 out of 10. The number is opaque — you don't know what produced it, so you can't reason about when to trust it and when to override it.

InDecision's factor breakdown is the opposite of opaque. When the API returns a 74.3% BEARISH conviction with the technical factor scoring NEUTRAL while the other four factors score BEARISH, you know exactly what the disagreement is. The technicals haven't confirmed what pattern, volume, timeframe alignment, and timing are all seeing. That's information you can act on — wait for technical confirmation, or size down because one factor is lagging.

This granularity is what makes the API useful for more than just "long or short" decisions. The per-factor signals tell you the quality of the conviction, not just the quantity. A 75% conviction where all five factors agree is a different signal than a 75% conviction where two factors are extremely bullish and three are neutral. The breakdown distinguishes between those scenarios.

How the Engine Works Behind the Endpoint

The Signal API is a thin layer over the InDecision engine — the same DualCaseAggregator that runs our live analysis. The engine builds two independent cases for every analysis: a bull case and a bear case. Each scorer evaluates the market data and contributes evidence to both sides. The winning case determines the bias. The spread between cases determines conviction.

Five weighted scorers contribute to the 100-point scale:

• Daily Pattern (30%) — candlestick pattern recognition across a 5-candle window, 14 pattern types
• Volume (25%) — volume anomaly detection, buyer/seller pressure from OHLC data
• Timeframe Alignment (20%) — multi-timeframe scoring across 1H, 4H, 1D, and 1W
• Technical (15%) — RSI(14), MACD(12,26,9), Bollinger Bands(20,2)
• Market Timing (10%) — session timing effects, day-of-week patterns

Post-scoring adjustment layers (mean reversion and cascade detection) modify the final output by up to 32 additional points in either direction. These layers catch edge cases the primary scorers miss — overextended moves ripe for reversal, or bottom-up confirmation signals propagating across timeframes.

The API doesn't simplify or round the engine output. What the aggregator produces is what the endpoint returns.

What This Does Not Do

The Signal API does not execute trades. It does not manage positions. It does not provide price targets, stop-loss levels, or entry timing. It is not a trading bot and should not be treated as one.

It is a conviction signal — a structured, quantitative assessment of directional bias with full factor transparency. What you build on top of that signal is up to you.

The gap between a conviction signal and a trading decision is where your edge lives. InDecision handles the analytical layer. The strategy layer — position sizing, risk management, entry logic — belongs to the builder consuming the API.

The framework's 67% accuracy over 7+ years of live trading data is the result of the full system: the signal, the conviction thresholds, and the discipline to ABSTAIN when the data is unclear. The API delivers the first two. The third is on you.