Tzivia-AI is an MCP server published by io.github.MMcBerry-bit, hosted at mcp.tzivia.dev. When last checked on 2026-11-22 it was working: it answers and lists its tools. It declares 10 tools, including tzivia_guide, tzivia_ask, tzivia_explain, tzivia_status, tzivia_synthesis.
Persistent memory for AI agents. Hosted cloud connector or local Windows install.
8 connections through 5 shared values. Tap any node.
Squares are this server's values; dots are other servers sharing them. A shared address or publisher is a lead, not proof of a common operator.
Inferred from the host name and headers: what a server runs on, not who runs it.
How to use Tzivia: what each tool does and when to use it, with an example. Call this when the user asks what Tzivia is, how it works, or which tool to use.
No parameters.
{
"name": "tzivia_guide",
"title": "How to use Tzivia",
"description": "How to use Tzivia: what each tool does and when to use it, with an example. Call this when the user asks what Tzivia is, how it works, or which tool to use.",
"inputSchema": {
"type": "object",
"properties": {}
},
"annotations": {
"title": "How to use Tzivia",
"readOnlyHint": true,
"openWorldHint": false,
"idempotentHint": true,
"destructiveHint": false
}
}Answer a plain-language question from what is stored in the user's Tzivia memory, returning the most relevant knowledge chains. Example: 'what did I decide about the launch date?'
| Parameter | Type | Description |
|---|---|---|
| top_k | integer | Max results to return |
| question * | string | Your question in plain language |
{
"name": "tzivia_ask",
"title": "Ask Tzivia",
"description": "Answer a plain-language question from what is stored in the user's Tzivia memory, returning the most relevant knowledge chains. Example: 'what did I decide about the launch date?'",
"inputSchema": {
"type": "object",
"required": [
"question"
],
"properties": {
"top_k": {
"type": "integer",
"default": 5,
"description": "Max results to return"
},
"question": {
"type": "string",
"description": "Your question in plain language"
}
}
},
"annotations": {
"title": "Ask Tzivia",
"readOnlyHint": true,
"openWorldHint": false,
"idempotentHint": true,
"destructiveHint": false
}
}Full knowledge breakdown for a concept: stored item data, which patterns it belongs to, which synthesis chains it appears in, and predictions.
| Parameter | Type | Description |
|---|---|---|
| value * | string | Concept to explain |
{
"name": "tzivia_explain",
"title": "Explain a concept",
"description": "Full knowledge breakdown for a concept: stored item data, which patterns it belongs to, which synthesis chains it appears in, and predictions.",
"inputSchema": {
"type": "object",
"required": [
"value"
],
"properties": {
"value": {
"type": "string",
"description": "Concept to explain"
}
}
},
"annotations": {
"title": "Explain a concept",
"readOnlyHint": true,
"openWorldHint": false,
"idempotentHint": true,
"destructiveHint": false
}
}Show a summary of the user's Tzivia memory: stored items, patterns, knowledge chains, learning cycles and current ranking weights.
No parameters.
{
"name": "tzivia_status",
"title": "Memory status",
"description": "Show a summary of the user's Tzivia memory: stored items, patterns, knowledge chains, learning cycles and current ranking weights.",
"inputSchema": {
"type": "object",
"properties": {}
},
"annotations": {
"title": "Memory status",
"readOnlyHint": true,
"openWorldHint": false,
"idempotentHint": true,
"destructiveHint": false
}
}List the knowledge chains (linked sequences of concepts) Tzivia has built from the user's saved data. Optionally filter by domain.
| Parameter | Type | Description |
|---|---|---|
| limit | integer | Max results |
| domain | string | Filter by domain (medical/financial/science/code/legal/general) |
{
"name": "tzivia_synthesis",
"title": "Knowledge chains",
"description": "List the knowledge chains (linked sequences of concepts) Tzivia has built from the user's saved data. Optionally filter by domain.",
"inputSchema": {
"type": "object",
"properties": {
"limit": {
"type": "integer",
"default": 20,
"description": "Max results"
},
"domain": {
"type": "string",
"default": "",
"description": "Filter by domain (medical/financial/science/code/legal/general)"
}
}
},
"annotations": {
"title": "Knowledge chains",
"readOnlyHint": true,
"openWorldHint": false,
"idempotentHint": true,
"destructiveHint": false
}
}Rate how useful the result of an earlier tzivia_recall was: score 0.0 = not useful, 0.5 = neutral, 1.0 = exactly right. Tzivia uses the rating to rank results better in future recalls.
| Parameter | Type | Description |
|---|---|---|
| query * | string | The same query string passed to tzivia_recall |
| score * | number | Quality score 0.0 (useless) to 1.0 (perfect) |
{
"name": "tzivia_feedback",
"title": "Rate a recall",
"description": "Rate how useful the result of an earlier tzivia_recall was: score 0.0 = not useful, 0.5 = neutral, 1.0 = exactly right. Tzivia uses the rating to rank results better in future recalls.",
"inputSchema": {
"type": "object",
"required": [
"query",
"score"
],
"properties": {
"query": {
"type": "string",
"description": "The same query string passed to tzivia_recall"
},
"score": {
"type": "number",
"description": "Quality score 0.0 (useless) to 1.0 (perfect)"
}
}
},
"annotations": {
"title": "Rate a recall",
"readOnlyHint": false,
"openWorldHint": false,
"idempotentHint": false,
"destructiveHint": false
}
}Save text to the user's Tzivia memory: notes, decisions, preferences, documents, or conversation excerpts the user wants kept. Any length is accepted; long input is processed in chunks. Returns a compress_id and a summary of what was stored. Use tzivia_recall to retrieve it later.
| Parameter | Type | Description |
|---|---|---|
| text * | string | Raw text to compress (any length -- streamed internally) |
| label | string | Human-readable label for this compression (optional) |
| cycles | integer | Learning cycles to run (0 = auto from corpus size) |
{
"name": "tzivia_compress",
"title": "Save to memory",
"description": "Save text to the user's Tzivia memory: notes, decisions, preferences, documents, or conversation excerpts the user wants kept. Any length is accepted; long input is processed in chunks. Returns a compress_id and a summary of what was stored. Use tzivia_recall to retrieve it later.",
"inputSchema": {
"type": "object",
"required": [
"text"
],
"properties": {
"text": {
"type": "string",
"description": "Raw text to compress (any length -- streamed internally)"
},
"label": {
"type": "string",
"default": "",
"description": "Human-readable label for this compression (optional)"
},
"cycles": {
"type": "integer",
"default": 0,
"description": "Learning cycles to run (0 = auto from corpus size)"
}
}
},
"annotations": {
"title": "Save to memory",
"readOnlyHint": false,
"openWorldHint": false,
"idempotentHint": false,
"destructiveHint": false
}
}Poll the status of an async ingestion job submitted via tzivia_compress when ingestion_mode is 'async'. Returns state (queued/running/done/failed), progress (0.0-1.0), chunks processed, tokens added, synthesis nodes built, and the full result once done.
| Parameter | Type | Description |
|---|---|---|
| job_id * | string | The job_id returned by tzivia_compress in async mode |
{
"name": "tzivia_job_status",
"title": "Save progress",
"description": "Poll the status of an async ingestion job submitted via tzivia_compress when ingestion_mode is 'async'. Returns state (queued/running/done/failed), progress (0.0-1.0), chunks processed, tokens added, synthesis nodes built, and the full result once done.",
"inputSchema": {
"type": "object",
"required": [
"job_id"
],
"properties": {
"job_id": {
"type": "string",
"description": "The job_id returned by tzivia_compress in async mode"
}
}
},
"annotations": {
"title": "Save progress",
"readOnlyHint": true,
"openWorldHint": false,
"idempotentHint": true,
"destructiveHint": false
}
}Search the user's Tzivia memory and return the most relevant saved information for a query, as a compact text summary of at most max_tokens. Pass compress_id from an earlier tzivia_compress call to search only that save, or leave it empty to search everything.
| Parameter | Type | Description |
|---|---|---|
| query * | string | The current question or topic to retrieve knowledge for |
| max_tokens | integer | Target size of the returned context string in tokens |
| compress_id | string | compress_id from tzivia_compress (empty = search all) |
| include_historical | boolean | Include facts superseded by a newer, contradicting update (Memory DNA). Default False = current view only; set True for 'what did we believe before X' queries. |
{
"name": "tzivia_recall",
"title": "Recall from memory",
"description": "Search the user's Tzivia memory and return the most relevant saved information for a query, as a compact text summary of at most max_tokens. Pass compress_id from an earlier tzivia_compress call to search only that save, or leave it empty to search everything.",
"inputSchema": {
"type": "object",
"required": [
"query"
],
"properties": {
"query": {
"type": "string",
"description": "The current question or topic to retrieve knowledge for"
},
"max_tokens": {
"type": "integer",
"default": 500,
"description": "Target size of the returned context string in tokens"
},
"compress_id": {
"type": "string",
"default": "",
"description": "compress_id from tzivia_compress (empty = search all)"
},
"include_historical": {
"type": "boolean",
"default": false,
"description": "Include facts superseded by a newer, contradicting update (Memory DNA). Default False = current view only; set True for 'what did we believe before X' queries."
}
}
},
"annotations": {
"title": "Recall from memory",
"readOnlyHint": true,
"openWorldHint": false,
"idempotentHint": true,
"destructiveHint": false
}
}Show where a stored belief came from. Given a node ID (from tzivia_synthesis or tzivia_recall) or a query, trace the evidence chain back to the original sources, with their labels, save times and trust scores.
| Parameter | Type | Description |
|---|---|---|
| query | string | Alternative to node_id -- trace the top recall match for this query |
| node_id | string | A specific node ID to trace (any tier) |
{
"name": "tzivia_provenance",
"title": "Why Tzivia believes this",
"description": "Show where a stored belief came from. Given a node ID (from tzivia_synthesis or tzivia_recall) or a query, trace the evidence chain back to the original sources, with their labels, save times and trust scores.",
"inputSchema": {
"type": "object",
"properties": {
"query": {
"type": "string",
"default": "",
"description": "Alternative to node_id -- trace the top recall match for this query"
},
"node_id": {
"type": "string",
"default": "",
"description": "A specific node ID to trace (any tier)"
}
}
},
"annotations": {
"title": "Why Tzivia believes this",
"readOnlyHint": true,
"openWorldHint": false,
"idempotentHint": true,
"destructiveHint": false
}
}Tzivia is the user's long-term memory, kept privately in their Tzivia Cloud account. Core tools -- tzivia_compress: Save a conversation, notes or a document into memory. Any size works; tzivia_recall: Pull back only what's relevant to a question, as a compact summary the AI can use; tzivia_ask: Ask a plain-language question about what Tzivia has learned; tzivia_status: See how much Tzivia knows: memories, patterns, learning cycles; tzivia_guide: Show this guide again. Save with tzivia_compress when the user asks to remember something or shares information they want kept; use tzivia_recall or tzivia_ask when an answer may depend on what the user saved earlier. tzivia_feedback rates how useful a recall was. If a tool result includes a 'tzivia_welcome' field, the user is new to Tzivia, and a short explanation of the core tools with one example may help; the full guide is available from tzivia_guide. When the user asks what Tzivia is or how to use it, call tzivia_guide.
Names, descriptions and schemas were written by the server and are shown as data.
Pattern matches on what the server declares. They show what it says it can do, not what it does.
None yet. Changes appear after a server's second daily check.