Effective 4 July 2026 · operated by Jiva Studio · contact support@jiva.studio
This policy covers the Listen to Sadhu MCP server at mcp.listentosadhu.app/mcp — the connector you add to an AI assistant. It is separate from the Listen to Sadhu app's policy: the MCP server has no accounts and a different, much smaller data flow.
A public, read-only MCP server over a corpus of Vedic scripture and recorded lectures. There is no sign-in, no user account, no cookies, and no payment — anyone's assistant may connect and read.
Only the tool arguments your assistant sends us to answer your request: search queries, verse or lecture references, and similar parameters. We do not receive your identity, your account, or your wider conversation with the assistant.
To run semantic search, the text of a search query is sent to a third-party, OpenAI-compatible embedding provider to turn it into a numeric vector; that provider processes the query under its own terms. Structured reads — verses, documents, lecture transcripts — are served from our own data and are not sent anywhere else.
We keep short-lived operational logs to run, secure and improve the service. A log entry for a tool call records the tool name, the query text and filters, the result count, timing, and a non-reversible hashed client identifier (a SHA-256 of the IP, truncated — the raw IP is never stored). Reverse-proxy access logs keep only a masked network prefix of the IP (the last octet removed), not the full address.
This is a public scriptural reference tool. Don't include personal or sensitive information in queries — as noted above, query text may be logged for a limited time.
We may update this policy; the effective date above reflects the current version. Material changes will be posted at this URL.
Questions? support@jiva.studio