Vedākṣha

MCP Integration

7 tools your agent already knows.

Vedākṣha exposes 7 typed tools via the Model Context Protocol (MCP). Every tool has a JSON schema — so any MCP-compatible agent can call them without custom prompting or output parsing.

Transport: JSON-RPC 2.0. Auth: OAuth 2.1. Errors are structured with machine-readable codes and self-correction hints so your agent can retry intelligently.

compute_natal_chart(julian_day, latitude, longitude, config?)

Computes a full natal chart and returns the complete ChartGraph JSON. Includes planets, house cusps, nakshatras, aspects, dignities, and yogas.

compute_dasha(chart_id, system?)

Returns the dasha tree for a previously computed chart. Supports Vimshottari, Yogini, and Chara dasha systems with up to 5 levels of sub-periods.

compute_vargas(chart_id, divisions[])

Computes one or more of the 16 Shodasha Varga divisional charts. Each varga is returned as an independent ChartGraph with its own planetary positions.

emit_graph(chart_id, format)

Emits the ChartGraph in the requested format. Supported formats: Cypher (Neo4j), SurrealQL (SurrealDB), JSON-LD, JSON, and EmbeddingText for RAG pipelines.

search_transits(planet, target, start_jd, end_jd, config?)

Streams exact transit moments in a date range. Returns TransitEvent objects with Julian Day, ingress/egress flag, and natural language description.

search_muhurta(criteria, start_jd, end_jd, location)

Finds auspicious time windows matching tithi, nakshatra, weekday, and hora criteria. Results are scored and sorted by combined auspiciousness.

describe_chart(chart_id, locale?)

Returns pre-written natural language descriptions for the chart's most significant features — yogas, planetary dignities, and dasha summary. Ready for agent relay.