A pasteable skeleton generated from the repository name. Check the project README for the exact package and arguments.
Star count
0 ★ · davidmosiah/astral-mcp · read from the GitHub search API · (checked 2026-10-16 20:00:20)
Source
collected from the punkpeye/awesome-mcp-servers list · type: MCP server
Harvest history
first seen 2026-10-09 · last seen 2026-10-16 · seen 32× · listed by 7 sources
Score
score 703 · 7 sources · mcp
FAQ (answers come from our records)
How do I install it?
We did not find an install command for this repository (our install coverage is only about 4%). Check its README or releases page.
Who is it for?
We filed it under these tags: MCP 服务. The tags come from the entry itself, not a taxonomy we imposed.
Who maintains it?
repository davidmosiah/astral-mcp;first catalogued 2026-10-09;last seen 2026-10-16。There is no official certification or quality rating here — we record sources and dates only.
Give your AI agent real astrology — natal charts, transits, synastry & moon phases. Computed locally, cross-checked by two independent ephemerides . No API key, no account. > ⚡ **Zero-setup install.** Wire it into Claude Desktop / Cursor / Hermes and call it immediately — no API key, no OAuth, no account: > npx -y astral-mcp **A local-first MCP server that turns birth data into a full, precision-audited astrological reading for AI agents.** Stateless and computational — nothing is stored, no credentials exist, and every tool but optional geocoding runs fully offline. Astral MCP is Delx's focused astrology product. See the Delx platform map for its boundary with the separate Protocol, Security, Wellness and Commerce properties. Built by David Mosiah. The astrology engine is ported from the Alkhemia app. New to the protocol pattern? Read **What Is an Astrology MCP Server?** for the separation between deterministic chart computation and model interpretation, the recommended agent workflow, and the precision/privacy boundaries. Ready to calculate one? Follow **How to Generate a Natal Chart with MCP** for the exact birthplace, timezone, unknown birth-time, payload and precision-audit workflow. Handling sensitive inputs? Read **How to Protect Birth Data in Astrology Agent Workflows** for payload minimization, local computation, the optional OpenStreetMap boundary and the limits of privacy_mode=summary. Why this exists Most astrology libraries are fragile single-engine wrappers, and most "astrology APIs" want a key and a subscription. Agents need something they can trust and call instantly. Astral MCP does two things differently: 1. **It just works.** npx -y astral-mcp and you're calling charts — no OAuth, no account, no birthplace database to install. 2. **It checks itself.** Every natal chart is computed with one ephemeris and then independently re-derived, planet by planet, with a second one. If they disagree beyond a tight tolerance, the chart is flagged review instead of silently returning a wrong placement. Setup in 60 seconds Add it to your MCP client (Claude Desktop, Cursor, Hermes, …): That's the whole setup. There is nothing to authenticate. Run it directly if you want: See it before you connect Call **astral_demo** for a fully-worked example chart (Greenwich, noon, Y2K) including its precision audit — no input, no network, no auth. It shows you the exact payload shape before you send real birth data. Try it with your agent > "What's my natal chart? I was born 23 Feb 1989, 14:30, in Fortaleza, Brazil." The agent calls astral_search_birthplace to resolve Fortaleza → lat/lon/timezone, then astral_compute_natal_chart. > "Any big transits hitting my chart this week?" → astral_current_transits > "How compatible are we?" (two birth datas) → astral_synastry > "What phase is the moon in today?" → astral_moon_phase Precision Astral MCP ships every natal chart with a precision audit. The primary engine (circular-natal-horoscope-js) computes the chart; the verifier (astronomy-engine) re-derives each planet's ecliptic longitude independently. A chart is verified only when every planet agrees within tolerance and lands in the same sign. Across a built-in accuracy suite of charts spanning 1945–2010 and six timezones, the **worst cross-engine disagreement is under 0.01°**. Run it yourself: