Agentic-India

Trijya Agent Cookbook — Data skill for Claude Code

Data community

India public data API and MCP server samples — search 3.7M MCA companies by name or CIN, 44M MSME Udyam registrations, PIN codes, IFSC bank branches, Central Acts and IPC to BNS mapping.

How to install Trijya Agent Cookbook

This entry records only its repository, not the path inside it, so there is no exact command to give. Open Agentic-India/trijya-agent-cookbook and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Trijya Agent Cookbook does

India public data API and MCP server samples — search 3.7M MCA companies by name or CIN, 44M MSME Udyam registrations, PIN codes, IFSC bank branches, Central Acts and IPC to BNS mapping. KYB and due diligence AI agents on Google ADK, Claude and Gemini.

Alternatives in Data

Private GPT — Complete API layer for private AI applications on local models: RAG, skills, tools, MCP, text-to-sql, and more 57.5k ★
MCP Server Chart — 🤖 A visualization mcp & skills contains 25+ visual charts using @antvis 4k ★
MCP Server Supabase — Supabase database and auth integration 2.6k ★

README

Trijya Cookbook — India public data API and MCP server samples

Working samples for [Trijya](https://trijya.in): search Indian companies by name or CIN, check MSME/Udyam registrations, resolve PIN codes and districts, look up IFSC bank branches, and map an old IPC section to its BNS replacement — from a REST API, from an MCP server, or from an AI agent you run yourself.

What Trijya is

India's public data is official, plentiful and almost unusable in a workflow. It is scattered across dozens of portals in different shapes, published as files and dashboards rather than as queries, keyed differently in every source so nothing joins cleanly, and it shows you only what is true today — when a register is updated, what it said last year is gone.

Trijya is that data as one layer: a REST API and an MCP server over India's company, enterprise, geography, banking, legal and statistical registers. Two things come with every answer.

**A date, always.** Each row carries what the register said (`knowledge_date`) and when Trijya read it (`capture_date`). Neither is "today". A government register refreshes on its own schedule, and anything you put in front of a user should say which day it speaks for.

**History, not just the present.** Trijya keeps versions rather than overwriting, so you can ask what the register said on a past date — whether a company was active on the day a contract was signed, not merely whether it is active now. That is the part a diligence or underwriting file actually needs, and it is the part no portal gives you.

Today it holds
Companies	3,740,426
MSME / Udyam enterprises	44,562,215
LLPs	514,494
Villages mapped	720,758
Bank branches	182,523
Local bodies	353,197
Health facilities	30,273
Central Acts	845

Plus macro and budget series, industrial classification codes, the current criminal-code sections with their mapping from the old ones, and a hashed archive of the source documents