Transparent Jyotish, exposed as MCP tools any AI agent can call — multi-school, 7 languages, every answer traceable to a classical source and a JPL-grade fact.
live from /stats · updated every 60s
Real output from get_dasha_timeline, rendered inline. Select a chart, school, and language to explore.
We own and ship every layer of the stack, all in production. So we move fast and respond — a change can flow from a raw ephemeris primitive to the engine to a shipping product in days, not quarters. No third-party ephemeris dependency, no vendor lock.
Ephemeris → engine → frontend — all ours, all live.
The explain tool returns the full derivation chain for any claim KundaliMCP makes — from prose interpretation back to an astronomical measurement.
Pass include=detail on any tool call, or call explain directly with a claim hash, to retrieve the full chain.
Each school is a sealed computation context. IDL runs once; Parāśarī, Jaimini, and KP interpretations run in parallel with no cross-contamination or silent defaults.
en · hi · sa · ta · te · kn · bn — built into the data layer, not layered on as i18n. Every yoga name, nakṣatra label, and narrative fragment carries all seven.
Every rule cites chapter and verse. No interpolation, no creative inference. Sources are BPHS, Sāravalī, Phaladīpikā, and Jaimini Sūtras.
Discover via tools/list — no SDK, no schema bundling, just standard MCP.
All tiers access every tool. Quota and key count scale with your tier.
Add mcp.kundalimcp.com/mcp to your agent's MCP config. Issue a key. Start calling tools.