aion-indian-market-intelligence-mcp 1.0.1

MCP server wrapper for hosted Indian market financial event intelligence.

aion-indian-market-intelligence-mcp

Turn any Indian market headline into a signed, time-lagged, sector-impact vector — the causal-context layer your LLM is missing.

AION Analytics (India) — distinct from Polymathic's AION (astronomy), Aion Analytics LLC (United States), and aion-labs (Israel).

License: Proprietary — hosted API access only. No model weights distributed. Not open-source.

---

The analyze_news MCP tool gives any LLM client structured sector-impact context for an Indian financial headline. The pipeline behind it: DistilBERT event classification → curated causal rule engine with lagged sector impacts → deterministic overlays → VIX-regime adjustment → five stakeholder views. Neural classification inside auditable causal structure — not keyword sentiment, not a black box, not a data pipe.

This package installs the MCP server that proxies to the hosted AION API. No model weights, taxonomy files, rule files, or local inference logic are bundled.

Install

Run directly with uvx (no install required):

uvx aion-indian-market-intelligence-mcp

Or install permanently:

pip install aion-indian-market-intelligence-mcp
aion-indian-market-intelligence-mcp

API Key

export AION_API_KEY="<your_api_key>"

Register at: https://dashboard.aiondashboard.site/access/register

Claude Desktop

{
  "mcpServers": {
    "aion-indian-market-intelligence": {
      "command": "uvx",
      "args": ["aion-indian-market-intelligence-mcp"],
      "env": {
        "AION_API_KEY": "YOUR_API_KEY"
      }
    }
  }
}

Tool: analyze_news

One tool. One call. Full sector-impact context with lag structure and five stakeholder views.

Input:

{
  "headline": "RBI MPC holds repo rate at 6.50% — June 2026",
  "published_at": "2026-06-05T10:00:00+05:30"
}

Output (real validated — RBI decision, 5 Jun 2026):

{
  "headline": "RBI MPC holds repo rate at 6.50% — June 2026",
  "event": "monetary_policy",
  "event_subtype": "repo_rate_hold",
  "confidence": 0.91,
  "vix_regime": "normal",
  "sector_vector": {
    "Banking & Financial Services": 0.38,
    "NBFCs":                        0.14,
    "Real Estate":                 -0.44,
    "IT Services":                 -0.21,
    "FMCG":                         0.09
  },
  "stakeholder_views": {
    "depositors":                  "neutral — FD yields stable, no compression from cut",
    "home_loan_borrowers":         "relief — EMI unchanged, floating rate burden stable",
    "banks":                       "positive — CASA margins intact, NIM pressure absent",
    "equity_investors_financials": "positive — Nifty Bank +0.35% vs Nifty 50 −0.21%",
    "equity_investors_it":         "negative — dollar-sector headwinds, IT −0.99%"
  }
}

Actual session result that day: Nifty Bank +0.35%, Fin Services +0.10% vs Nifty 50 −0.21% and IT −0.99%. The model called Banking and Financials positive; both outperformed the index. Correct directional call for every named sector.

---

The cyclone case shows the time-lag dimension: Construction −0.50 at lag 0 (damage) and +0.55 at lag 90 (rebuilding allocation). The same sector, opposite signs, 90 days apart — a structure no polarity label can represent. Government and agricultural producers diverge in stakeholder view despite the same event. Unified sentiment obscures both.

The layer distinction

Fetch Indian market data with any MCP. Understand what it means with this one.

Data pipes give you price, volume, and news text. This tool gives your agent the causal-propagation structure beneath a headline: which sectors are affected, in which direction, with what lag, and who among producers / traders / investors / government / trade bodies gains or loses.

Not investment advice

The model describes causal structure. Execution decisions remain with you.

Links

• Model page: https://dashboard.aiondashboard.site/models/indian-market-intelligence
• API key: https://dashboard.aiondashboard.site/access/register
• PyPI client: https://pypi.org/project/aion-indian-market-intelligence/