mindgraph-mcp 0.9.0 → 0.10.0

package/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2024 Shuruheel Ahamed
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

package/README.md

@@ -62,6 +62,8 @@ Or drop this into your Claude Desktop config by hand (`~/Library/Application Sup
 | `mindgraph_synthesize` | Project-scoped cross-document synthesis — mine signals, spawn Article-generation jobs |
 | `mindgraph_ontology` | Operational Ontology (Layer 7) — typed domain objects, NL queries with provenance, extraction proposal review (approve/reject), object linking |
 
+Plus **dynamically generated read tools**: when the tenant's active schema declares object types, the server renders per-type `search_<objs>` / `get_<obj>` / `summarize_<obj>` tools (from the cloud's `GET /v1/ontology/tools` manifest) at session start. `summarize_<obj>` returns the object — whether mapped from a connected SQL database or extracted from documents — *plus* its cognitive context (claims, risks, decisions, evidence).
+
 **7 prompt templates** for common workflows:
 
 `graph-summary`, `active-goals`, `open-questions`, `review-contradictions`, `knowledge-about`, `daily-briefing`, `capture-conversation`.

package/dist/cli.js

@@ -1011,7 +1011,7 @@ var init_tools = __esm({
       },
       {
         name: "mindgraph_retrieve",
-        description: "Search and explore the knowledge graph. Use 'context' (default) for BM25 keyword retrieval \u2014 returns graph nodes only by default (labels, summaries, types, confidence); set include_chunks=true to also get full source text. Pass 1-3 discriminating keywords, not sentences. GOOD: 'Kissinger NATO'. BAD: 'What is Kissinger\\'s view on NATO?'. Use 'document_index' to list all ingested documents (titles, dates, UIDs) for orientation. Use 'semantic' when keywords return nothing \u2014 conceptual/fuzzy queries. Use 'hybrid' for keyword + semantic. Use 'text' for fast keyword-only. Use 'neighborhood'/'chain'/'path'/'subgraph' for graph traversal. Use 'preferences' for advice or recommendation requests ('suggest a hotel', 'what should I read?') \u2014 it returns the user's stated/learned preferences relevant to the query (pass the topic as 'query'), so your answer reflects what they actually like instead of being generic. Other structured actions ('active_goals', 'open_questions', etc.) only when the user explicitly asks.",
+        description: "Search and explore the knowledge graph. Use 'context' (default) for BM25 keyword retrieval \u2014 returns graph nodes only by default (labels, summaries, types, confidence); set include_chunks=true to also get full source text. Pass 1-3 discriminating keywords, not sentences. GOOD: 'Kissinger NATO'. BAD: 'What is Kissinger\\'s view on NATO?'. Use 'document_index' to list all ingested documents (titles, dates, UIDs) for orientation. Use 'semantic' when keywords return nothing \u2014 conceptual/fuzzy queries. Use 'hybrid' for keyword + semantic. Use 'text' for fast keyword-only. Use 'neighborhood'/'chain'/'path'/'subgraph' for graph traversal \u2014 traversal steps carry path_cost (lower = stronger connection; sum of -ln(edge weight)) and path_confidence (product of edge confidences; a ranking signal, not a calibrated probability) for the returned path, useful for ranking results; they score the path returned, not the best possible path. Use 'preferences' for advice or recommendation requests ('suggest a hotel', 'what should I read?') \u2014 it returns the user's stated/learned preferences relevant to the query (pass the topic as 'query'), so your answer reflects what they actually like instead of being generic. Other structured actions ('active_goals', 'open_questions', etc.) only when the user explicitly asks.",
         inputSchema: {
           type: "object",
           properties: {
@@ -1311,6 +1311,84 @@ var init_tools = __esm({
   }
 });
 
+// src/generated-tools.ts
+async function getGeneratedTools(client2) {
+  if (cache && Date.now() - cache.fetchedAt < CACHE_TTL_MS) {
+    return { tools: cache.tools, byName: cache.byName };
+  }
+  try {
+    const c = client2;
+    if (typeof c.listOntologyTools !== "function") {
+      cache = { tools: [], byName: /* @__PURE__ */ new Map(), fetchedAt: Date.now() };
+      return { tools: cache.tools, byName: cache.byName };
+    }
+    const resp = await c.listOntologyTools();
+    const byName = /* @__PURE__ */ new Map();
+    const tools = [];
+    for (const d of resp.tools ?? []) {
+      byName.set(d.name, d);
+      tools.push({
+        name: d.name,
+        description: d.description,
+        inputSchema: d.input_schema
+      });
+    }
+    cache = { tools, byName, fetchedAt: Date.now() };
+    return { tools, byName };
+  } catch {
+    cache = { tools: [], byName: /* @__PURE__ */ new Map(), fetchedAt: Date.now() };
+    return { tools: cache.tools, byName: cache.byName };
+  }
+}
+async function handleGeneratedTool(client2, desc, args) {
+  try {
+    switch (desc.maps_to) {
+      case "search": {
+        const query = args.query;
+        if (!query) return err2(`query is required for ${desc.name}`);
+        const limit = typeof args.limit === "number" ? args.limit : void 0;
+        return ok2(
+          await client2.searchDomainObjects(query, {
+            schema_id: desc.schema_id,
+            object_types: [desc.object_type],
+            limit
+          })
+        );
+      }
+      case "object": {
+        const uid = args.uid;
+        if (!uid) return err2(`uid is required for ${desc.name}`);
+        return ok2(await client2.getDomainObject(uid));
+      }
+      case "object_context": {
+        const uid = args.uid;
+        if (!uid) return err2(`uid is required for ${desc.name}`);
+        const depth = typeof args.depth === "number" ? args.depth : void 0;
+        return ok2(await client2.getDomainObjectContext(uid, depth));
+      }
+      default:
+        return err2(`unknown generated tool mapping: ${desc.maps_to}`);
+    }
+  } catch (e) {
+    return err2(e instanceof Error ? e.message : String(e));
+  }
+}
+var ok2, err2, CACHE_TTL_MS, cache;
+var init_generated_tools = __esm({
+  "src/generated-tools.ts"() {
+    "use strict";
+    ok2 = (data) => ({
+      content: [{ type: "text", text: JSON.stringify(data, null, 2) }]
+    });
+    err2 = (message) => ({
+      content: [{ type: "text", text: JSON.stringify({ error: message }) }],
+      isError: true
+    });
+    CACHE_TTL_MS = 5 * 60 * 1e3;
+    cache = null;
+  }
+});
+
 // src/index.ts
 var src_exports = {};
 async function main() {
@@ -1329,6 +1407,7 @@ var init_src = __esm({
     import_types = require("@modelcontextprotocol/sdk/types.js");
     import_mindgraph = require("mindgraph");
     init_tools();
+    init_generated_tools();
     API_KEY = process.env.MINDGRAPH_API_KEY;
     BASE_URL = process.env.MINDGRAPH_BASE_URL || "https://api.mindgraph.cloud";
     AGENT_ID = process.env.MINDGRAPH_AGENT_ID || "mcp";
@@ -1422,7 +1501,7 @@ Capture knowledge when the user shares something worth remembering:
     server = new import_server.Server(
       {
         name: "mindgraph",
-        version: "0.7.1"
+        version: "0.9.1"
       },
       {
         capabilities: {
@@ -1434,7 +1513,8 @@ Capture knowledge when the user shares something worth remembering:
       }
     );
     server.setRequestHandler(import_types.ListToolsRequestSchema, async () => {
-      return { tools: TOOLS };
+      const generated = await getGeneratedTools(client);
+      return { tools: [...TOOLS, ...generated.tools] };
     });
     server.setRequestHandler(import_types.CallToolRequestSchema, async (request) => {
       const { name, arguments: args } = request.params;
@@ -1442,6 +1522,11 @@ Capture knowledge when the user shares something worth remembering:
       if (!toolArgs.agent_id) {
         toolArgs.agent_id = AGENT_ID;
       }
+      const generated = await getGeneratedTools(client);
+      const desc = generated.byName.get(name);
+      if (desc) {
+        return handleGeneratedTool(client, desc, toolArgs);
+      }
       return handleTool(client, name, toolArgs);
     });
     PROMPTS = [

package/dist/index.js

@@ -297,7 +297,7 @@ var TOOLS = [
   },
   {
     name: "mindgraph_retrieve",
-    description: "Search and explore the knowledge graph. Use 'context' (default) for BM25 keyword retrieval \u2014 returns graph nodes only by default (labels, summaries, types, confidence); set include_chunks=true to also get full source text. Pass 1-3 discriminating keywords, not sentences. GOOD: 'Kissinger NATO'. BAD: 'What is Kissinger\\'s view on NATO?'. Use 'document_index' to list all ingested documents (titles, dates, UIDs) for orientation. Use 'semantic' when keywords return nothing \u2014 conceptual/fuzzy queries. Use 'hybrid' for keyword + semantic. Use 'text' for fast keyword-only. Use 'neighborhood'/'chain'/'path'/'subgraph' for graph traversal. Use 'preferences' for advice or recommendation requests ('suggest a hotel', 'what should I read?') \u2014 it returns the user's stated/learned preferences relevant to the query (pass the topic as 'query'), so your answer reflects what they actually like instead of being generic. Other structured actions ('active_goals', 'open_questions', etc.) only when the user explicitly asks.",
+    description: "Search and explore the knowledge graph. Use 'context' (default) for BM25 keyword retrieval \u2014 returns graph nodes only by default (labels, summaries, types, confidence); set include_chunks=true to also get full source text. Pass 1-3 discriminating keywords, not sentences. GOOD: 'Kissinger NATO'. BAD: 'What is Kissinger\\'s view on NATO?'. Use 'document_index' to list all ingested documents (titles, dates, UIDs) for orientation. Use 'semantic' when keywords return nothing \u2014 conceptual/fuzzy queries. Use 'hybrid' for keyword + semantic. Use 'text' for fast keyword-only. Use 'neighborhood'/'chain'/'path'/'subgraph' for graph traversal \u2014 traversal steps carry path_cost (lower = stronger connection; sum of -ln(edge weight)) and path_confidence (product of edge confidences; a ranking signal, not a calibrated probability) for the returned path, useful for ranking results; they score the path returned, not the best possible path. Use 'preferences' for advice or recommendation requests ('suggest a hotel', 'what should I read?') \u2014 it returns the user's stated/learned preferences relevant to the query (pass the topic as 'query'), so your answer reflects what they actually like instead of being generic. Other structured actions ('active_goals', 'open_questions', etc.) only when the user explicitly asks.",
     inputSchema: {
       type: "object",
       properties: {
@@ -1285,6 +1285,78 @@ async function handleOntology(client2, args) {
   }
 }
 
+// src/generated-tools.ts
+var ok2 = (data) => ({
+  content: [{ type: "text", text: JSON.stringify(data, null, 2) }]
+});
+var err2 = (message) => ({
+  content: [{ type: "text", text: JSON.stringify({ error: message }) }],
+  isError: true
+});
+var CACHE_TTL_MS = 5 * 60 * 1e3;
+var cache = null;
+async function getGeneratedTools(client2) {
+  if (cache && Date.now() - cache.fetchedAt < CACHE_TTL_MS) {
+    return { tools: cache.tools, byName: cache.byName };
+  }
+  try {
+    const c = client2;
+    if (typeof c.listOntologyTools !== "function") {
+      cache = { tools: [], byName: /* @__PURE__ */ new Map(), fetchedAt: Date.now() };
+      return { tools: cache.tools, byName: cache.byName };
+    }
+    const resp = await c.listOntologyTools();
+    const byName = /* @__PURE__ */ new Map();
+    const tools = [];
+    for (const d of resp.tools ?? []) {
+      byName.set(d.name, d);
+      tools.push({
+        name: d.name,
+        description: d.description,
+        inputSchema: d.input_schema
+      });
+    }
+    cache = { tools, byName, fetchedAt: Date.now() };
+    return { tools, byName };
+  } catch {
+    cache = { tools: [], byName: /* @__PURE__ */ new Map(), fetchedAt: Date.now() };
+    return { tools: cache.tools, byName: cache.byName };
+  }
+}
+async function handleGeneratedTool(client2, desc, args) {
+  try {
+    switch (desc.maps_to) {
+      case "search": {
+        const query = args.query;
+        if (!query) return err2(`query is required for ${desc.name}`);
+        const limit = typeof args.limit === "number" ? args.limit : void 0;
+        return ok2(
+          await client2.searchDomainObjects(query, {
+            schema_id: desc.schema_id,
+            object_types: [desc.object_type],
+            limit
+          })
+        );
+      }
+      case "object": {
+        const uid = args.uid;
+        if (!uid) return err2(`uid is required for ${desc.name}`);
+        return ok2(await client2.getDomainObject(uid));
+      }
+      case "object_context": {
+        const uid = args.uid;
+        if (!uid) return err2(`uid is required for ${desc.name}`);
+        const depth = typeof args.depth === "number" ? args.depth : void 0;
+        return ok2(await client2.getDomainObjectContext(uid, depth));
+      }
+      default:
+        return err2(`unknown generated tool mapping: ${desc.maps_to}`);
+    }
+  } catch (e) {
+    return err2(e instanceof Error ? e.message : String(e));
+  }
+}
+
 // src/index.ts
 var API_KEY = process.env.MINDGRAPH_API_KEY;
 var BASE_URL = process.env.MINDGRAPH_BASE_URL || "https://api.mindgraph.cloud";
@@ -1379,7 +1451,7 @@ Capture knowledge when the user shares something worth remembering:
 var server = new import_server.Server(
   {
     name: "mindgraph",
-    version: "0.7.1"
+    version: "0.9.1"
   },
   {
     capabilities: {
@@ -1391,7 +1463,8 @@ var server = new import_server.Server(
   }
 );
 server.setRequestHandler(import_types.ListToolsRequestSchema, async () => {
-  return { tools: TOOLS };
+  const generated = await getGeneratedTools(client);
+  return { tools: [...TOOLS, ...generated.tools] };
 });
 server.setRequestHandler(import_types.CallToolRequestSchema, async (request) => {
   const { name, arguments: args } = request.params;
@@ -1399,6 +1472,11 @@ server.setRequestHandler(import_types.CallToolRequestSchema, async (request) =>
   if (!toolArgs.agent_id) {
     toolArgs.agent_id = AGENT_ID;
   }
+  const generated = await getGeneratedTools(client);
+  const desc = generated.byName.get(name);
+  if (desc) {
+    return handleGeneratedTool(client, desc, toolArgs);
+  }
   return handleTool(client, name, toolArgs);
 });
 var PROMPTS = [

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "mindgraph-mcp",
-  "version": "0.9.0",
+  "version": "0.10.0",
   "description": "MCP server for MindGraph — connect Claude Desktop, Claude Code, and any MCP client to your knowledge graph",
   "main": "dist/index.js",
   "types": "dist/index.d.ts",
@@ -13,6 +13,7 @@
   "scripts": {
     "build": "tsup",
     "lint": "tsc --noEmit",
+    "test": "vitest run",
     "dev": "tsup --watch"
   },
   "keywords": [
@@ -27,11 +28,12 @@
   "license": "MIT",
   "dependencies": {
     "@modelcontextprotocol/sdk": "^1.12.1",
-    "mindgraph": "^0.5.0"
+    "mindgraph": "^0.11.0"
   },
   "devDependencies": {
     "@types/node": "^22.0.0",
     "tsup": "^8.0.0",
-    "typescript": "^5.7.0"
+    "typescript": "^5.7.0",
+    "vitest": "^3.0.0"
   }
 }