Celestial computation.
Agentic precision.

The astronomical engine built from the ground up for AI agents. Sub-arcsecond planetary positions. Vedic astrology in the type system. Charts as property graphs.

$cargo add vedaksha

Get Started Why AI-First

AI-First

Built for agents, not adapted.

Pure functions. Zero state. No initialization. Your AI agent calls one function and gets a complete chart back — typed enums, not integer codes.

7 MCP tools with JSON schemas

Structured errors with self-correction hints

Streaming transit search

PII-blind — no personal data touches the engine

Vedic-First

Jyotish in the type system.

27 nakshatras, 50 yogas, 3 dasha systems, and 16 divisional charts — all with 7-language localization in English, Hindi, Sanskrit, Tamil, Telugu, Kannada, and Bengali.

Vimshottari, Yogini, and Chara dashas

Complete Shadbala (all 6 components)

44 ayanamsha systems

Muhurta search with tithi scoring

Graph-Native

Charts are graphs. Finally.

Every chart is a property graph with 10 node types and 13 edge types. Emit directly to Neo4j, SurrealDB, or JSON-LD. Feed into RAG pipelines for semantic search.

Deterministic IDs — same input, same graph

Cypher, SurrealQL, JSON-LD emitters

Embedding text for vector stores

Data classification for GDPR compliance

Five lines of Rust

Your first chart.

Pass a date, latitude, and longitude. Get back planetary positions, house cusps, nakshatras, aspects, dignities, and yogas — all typed, all in one call.

Rustcargo add vedaksha

Pythonpip install vedaksha

WASMnpm add vedaksha-wasm

main.rs

use vedaksha::prelude::*;

let jd = calendar_to_jd(2024, 3, 20, 12.0);
let chart = compute_chart(
  jd, 28.6139, 77.2090,
  &ChartConfig::vedic()
);

for planet in &chart.planets {
  println!("{}: {}° {}",
    planet.name, planet.longitude,
    planet.sign);
}

Sun: 359.8° Pisces
Moon: 127.5° Leo
Mars: 309.1° Aquarius
Jupiter: 40.2° Taurus

619automated tests

3platforms

212public API items

15+cited sources

What's inside

Everything you need. Nothing you don't.

Planetary Engine

Sub-arcsecond positions from NASA JPL DE440/441. Full IAU precession, nutation, aberration pipeline.

Vedic Astrology

27 nakshatras, 3 dasha systems, 50 yogas, 16 vargas, complete Shadbala — not a plugin.

House Systems

Placidus, Koch, Whole Sign, Equal, Campanus, Regiomontanus, Porphyry, Morinus, Alcabitius, Sripathi.

MCP Server

7 typed tools for AI agents. JSON-RPC 2.0 with structured errors and self-correction hints.

Graph Emitters

Output to Neo4j Cypher, SurrealDB, JSON-LD, or RAG-optimized text for vector embedding.

44 Ayanamsha

Every major sidereal tradition — Lahiri, Fagan-Bradley, Krishnamurti, Aryabhata, and 40 more.

Runs Everywhere

Native Rust, WebAssembly for browsers, Python via PyO3. One codebase, three targets.

7 Languages

Every planet, sign, nakshatra, and yoga name in English, Hindi, Sanskrit, Tamil, Telugu, Kannada, Bengali.

Zero Legacy

Built from scratch using NASA and academic sources. No borrowed code. Free for non-commercial use.

For AI builders

Your agent calls. Gets a chart.

Connect any MCP-compatible AI agent — Claude, GPT, or custom. JSON schemas mean your agent already knows the input format. Structured errors mean it can self-correct.

compute_natal_chart— Full chart → ChartGraph JSON

compute_dasha— Dasha tree with 5 levels of sub-periods

compute_vargas— Any of 16 divisional charts

emit_graph— ChartGraph → Cypher / SurrealQL / JSON-LD

search_transits— Find exact transit moments in a date range

search_muhurta— Find auspicious times by nakshatra and tithi

Start computing.

Free for non-commercial use. $500 one-time for commercial.

Get Started Pricing

All implementations must display attribution

Celestial computation.Agentic precision.