Astral MCP
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
No arguments | |||
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": true
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| astral_compute_natal_chartA | Compute a full natal (birth) chart from birth data: planet signs/degrees/houses, retrogrades, major aspects with orb and strength, the Ascendant/MC, and a derived signature (dominant element/modality, chart pattern, stelliums, angular planets). By default the chart is precision-audited by a second independent ephemeris. This is the primary tool. If you only have a city name, call astral_search_birthplace first to get latitude/longitude/timezone. Use privacy_mode to control payload size (a full chart is large): 'full' (default) returns everything; 'structured' drops redundant fields; 'summary' returns only the luminaries, Ascendant, chart signature and top aspects to save tokens. |
| astral_current_transitsA | Read the current (or a chosen date's) planetary transits against a natal chart. Returns the active transit aspects to the natal planets and angles, upcoming activation windows over the next days, and the current moon phase. Supply the same birth data you would for a natal chart; pass on_date/on_time to read a specific moment. Useful for 'what's happening for me astrologically' questions. Set privacy_mode=summary for just the moon phase and top transit aspects (omitting the full current-sky planet map), 'structured' for a leaner payload, or 'full' (default) for everything. |
| astral_synastryA | Compare two birth charts (synastry). Returns the inter-chart aspects between the two people's planets and angles, plus scored dimensions (harmony, chemistry, communication, growth) and an overall 0–100 score with a tone. Provide both people's birth data under 'person' and 'partner'. Resolve any city names with astral_search_birthplace first. Set privacy_mode=summary for just the score, tone, dimensions and strongest aspects; 'full' (default) returns the complete aspect list. |
| astral_moon_phaseA | Get the moon phase (new, waxing, full, waning, etc.), the moon's sign, illumination percentage and sun–moon angle for now or a chosen local date/time. Location-independent — no birth data needed. Good for ritual timing, journaling prompts, or 'what phase is the moon in' questions. |
| astral_search_birthplaceA | Geocode a city or place name into latitude, longitude and IANA timezone using OpenStreetMap (free, no key). Call this FIRST whenever you only have a place name, then pass the chosen match's coordinates and timezone to the chart tools. Returns several matches; pick the one whose displayName matches the intended birthplace. |
| astral_demoA | Return a fully-worked example natal chart (Greenwich, noon, 2000-01-01) including a precision audit, so you can see the exact payload shape before sending real birth data. The returned 'input' is a ready-to-send argument object for astral_compute_natal_chart — copy it and swap in real birth data; 'chart' is exactly what that call returns at the default privacy_mode=full. No input, no network, no auth. |
| astral_capabilitiesA | Explain what astral-mcp can compute: supported house systems, zodiacs, bodies, aspects and features, the dual-engine precision model, what is NOT included (nodes, Chiron, interpretations), and the recommended agent workflow. Read this first to plan a sequence of calls. |
| astral_data_inventoryA | Inventory the data domains this server exposes (natal, transits, relationship, geocoding), which tools serve each, and the recommended first calls. Does not compute anything or take any input beyond response_format. |
| astral_agent_manifestA | Agent-oriented setup and usage guide: install command, recommended first calls, the full tool list, agent rules (resolve birthplace first, pass birthplace timezone, birth_time caveats) and troubleshooting. Call this to learn how to drive astral-mcp correctly. |
| astral_connection_statusA | Health check. The server is stateless and local, so this verifies the Node version and proves both bundled ephemerides load and agree by computing a sample chart and running the dual-engine precision audit. Call this to confirm the install is working before computing real charts. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 10 tools
The four computational tools (natal chart, transits, synastry, moon phase) are clearly distinct, but the three informational tools (capabilities, data inventory, agent manifest) overlap in purpose and could confuse an agent choosing which to call. The remaining support tools (birthplace search, demo, connection status) are well-separated.
All tools share the consistent astral_ prefix and use snake_case, giving the set a unified look. However, naming style is mixed: some use verb_noun (compute_natal_chart, search_birthplace) while most are plain nouns (synastry, moon_phase, demo, capabilities), which is a minor inconsistency.
Ten tools is a reasonable count for an astrology server, and each core capability has a dedicated tool. However, five of the ten are meta/informational tools (capabilities, inventory, manifest, demo, status), which feels slightly heavy relative to the five functional tools, though still within an acceptable range.
The server covers the main astrological workflows: natal charts, transits, synastry, moon phase, and geocoding prerequisites. Missing advanced features like progressions or solar returns are acknowledged omissions rather than dead ends, so agents can complete typical tasks without gaps.