Bloomberg-grade prediction-market intelligence, on-demand.

Authentication

Every request carries an API token in the Authorization header:

curl -X POST https://kairon.trade/api/kaironos/intel/market.macro_snapshot \
  -H "Authorization: Bearer kos_..." \
  -H "Content-Type: application/json" \
  -d '{ "category": "recession", "limit": 10 }'

Issue tokens at /account/api-tokens. Plaintext is shown once; the hash is stored. Revoke any token instantly.

Quotas & rate limits

Tied to the caller's KaironOS plan (separate from KaironOS task quota — intel queries are reads, tasks are writes). Overage is $0.05 per intel query above quota.

429 responses include Retry-After in seconds.

market.macro_snapshot

Macro Snapshot

Top prediction-market signals across recession / fed / geopolitical / crypto / election with current YES probability + 7-day delta.

Request

POST /api/kaironos/intel/market.macro_snapshot
Authorization: Bearer kos_...
Content-Type: application/json

{
  "date": "2026-05-16",
  "category": "recession",
  "limit": 10
}

Response shape

{
  "generated_at": "ISO timestamp",
  "rows": [
    {
      "question": "string",
      "category": "recession | fed | geopolitical | crypto | election",
      "current_prob": "0..1",
      "delta_7d": "-1..1 (price change vs 7d ago)",
      "volume_usd": "number",
      "source": "polymarket"
    }
  ]
}

anomaly.detect

Anomaly Detect

Whale-pattern anomaly detector applied to caller-supplied time series. Confidence calibrated against Kairon whale-tracker accuracy.

Request

POST /api/kaironos/intel/anomaly.detect
Authorization: Bearer kos_...
Content-Type: application/json

{
  "series": [
    {
      "t": "2026-05-09T00:00:00Z",
      "v": 100
    },
    {
      "t": "2026-05-10T00:00:00Z",
      "v": 104
    },
    {
      "t": "2026-05-11T00:00:00Z",
      "v": 142
    },
    {
      "t": "2026-05-12T00:00:00Z",
      "v": 108
    }
  ],
  "sensitivity": 0.7,
  "vertical": "devops"
}

Response shape

{
  "hits": [
    {
      "t": "ISO",
      "v": "number",
      "z_score": "number",
      "confidence": "0..1",
      "reason": "string"
    }
  ],
  "baseline": {
    "mean": "number",
    "stddev": "number"
  },
  "calibration_note": "string"
}

news.semantic_search

News Semantic Search

News ranked by downstream prediction-market reaction magnitude — not engagement or SEO. pgvector cosine over 768-dim nomic-embed-text embeddings; falls back to ILIKE.

Request

POST /api/kaironos/intel/news.semantic_search
Authorization: Bearer kos_...
Content-Type: application/json

{
  "query": "fed rate cut",
  "recency_days": 30,
  "limit": 10,
  "min_market_impact": 0.2
}

Response shape

{
  "hits": [
    {
      "id": "uuid",
      "title": "string",
      "url": "https://kairon.trade/media/...",
      "published_at": "ISO",
      "source": "string",
      "market_impact": "0..1",
      "related_markets": [
        "string"
      ],
      "summary": "string"
    }
  ],
  "ranking_note": "string"
}

forecast.calibrated

Forecast (Calibrated)

Probability + 5/95 confidence band derived from Kairon Score calibration over resolved Polymarket markets. Honest about basis: includes sample size + Brier score.

Request

POST /api/kaironos/intel/forecast.calibrated
Authorization: Bearer kos_...
Content-Type: application/json

{
  "event_description": "US Fed cuts rates by 50bp at next FOMC",
  "horizon_days": 45
}

Response shape

{
  "probability": "0..1",
  "confidence_low": "0..1",
  "confidence_high": "0..1",
  "comparable_events": [
    {
      "description": "string",
      "outcome": "boolean",
      "days_to_resolution": "integer"
    }
  ],
  "calibration_basis": "string (states sample size + Brier score)"
}

Code samples

Node / TypeScript

const r = await fetch('https://kairon.trade/api/kaironos/intel/market.macro_snapshot', {
  method: 'POST',
  headers: {
    Authorization: `Bearer ${process.env.KAIRON_API_TOKEN}`,
    'Content-Type': 'application/json',
  },
  body: JSON.stringify({ category: 'recession', limit: 10 }),
})
if (!r.ok) throw new Error(`Kairon API ${r.status}`)
const { rows } = await r.json()

Python

import os, requests
r = requests.post(
  "https://kairon.trade/api/kaironos/intel/market.macro_snapshot",
  headers={"Authorization": f"Bearer {os.environ['KAIRON_API_TOKEN']}"},
  json={"category": "recession", "limit": 10},
)
r.raise_for_status()
rows = r.json()["rows"]

Claude Desktop / Cursor (MCP)

Use the official Kairon MCP server — no manual HTTP wiring. See mcp-server/README.md in the source repo.

Compliance & disclaimers

Every intel-tool response carries a top-level disclaimer string. Display it alongside any user-facing rendering of the result; do not strip it in proxies.

“Kairon Intelligence outputs are informational only and do not constitute investment, legal, or trading advice. Prediction-market data reflects bettor positioning, not endorsements or guarantees. Past calibration does not predict future accuracy.”

• Audit trail: every paid action writes to guardian_logs, immutable for 5 years (DB trigger refuses earlier DELETE and row mutation).
• SIEM export: available to Enterprise as NDJSON / CSV streaming download.
• Data provenance: Polymarket gamma-api (markets), nomic-embed-text 768-dim (news), historical Score curves (forecast).

Error responses