@decantr/mcp-server 3.6.1 → 3.7.0

package/README.md

@@ -12,7 +12,7 @@ AI Frontend Governance for codebases touched by AI agents. Give Claude, Cursor,
 
 ![Decantr MCP demo](https://raw.githubusercontent.com/decantr-ai/decantr/main/packages/mcp-server/assets/decantr-demo.gif)
 
-- **Structured Contract context** -- gives your AI assistant patterns, layouts, component specs, typed graph context, Brownfield/Hybrid authority, local law, behavior obligations, and task-time context instead of letting it guess
+- **Structured Contract context** -- gives your AI assistant patterns, layouts, component specs, shared Brownfield discovery metadata, typed graph context, Brownfield/Hybrid authority, local law, behavior obligations, and task-time context instead of letting it guess
 - **Evidence-backed repair loops** -- gives AI agents Project Health, component reuse drift, accepted behavior-obligation drift, accepted style bridge drift, stable diagnostic codes, typed repair IDs, graph-anchored Evidence Bundles, workspace health, and scoped repair prompts without uploading source
 - **Drift detection** -- catches when generated code deviates from your design intent
 - **Zero config** -- run with `npx`, no API keys or accounts required
@@ -38,20 +38,32 @@ Restart Claude Desktop. The Decantr tools will appear automatically.
 
 ### Cursor
 
-Create `.cursor/mcp.json` in your project root:
+Preferred setup:
+
+```bash
+npx @decantr/cli connect cursor
+```
+
+From a monorepo root:
+
+```bash
+pnpm exec decantr connect cursor --project apps/web
+```
+
+The connector preserves existing MCP servers and writes the project rule that tells Cursor Agent to call Decantr task context before route edits. For manual setup, create `.cursor/mcp.json` in your project root:
 
 ```json
 {
   "mcpServers": {
     "decantr": {
       "command": "npx",
-      "args": ["@decantr/mcp-server"]
+      "args": ["-y", "@decantr/mcp-server"]
     }
   }
 }
 ```
 
-Restart Cursor. The tools are available in Agent mode.
+Restart Cursor. The tools are available in Agent mode. Add a project rule equivalent to `decantr connect cursor` if you configure MCP manually.
 
 ### Windsurf
 
@@ -126,12 +138,12 @@ The server exposes a hard 8-tool MCP surface. Pass an `action` to select the rou
 
 | Tool | Description | Example Input |
 |------|-------------|---------------|
-| `decantr_project` | Read local project state and workspace health | `{ "action": "state", "project_path": "apps/web" }` |
+| `decantr_project` | Read local project state, shared discovery summary, and workspace health | `{ "action": "state", "project_path": "apps/web" }` |
 | `decantr_contract` | Read Essence, validate, check drift, generate a skeleton, or read the Contract capsule | `{ "action": "validate", "path": "./decantr.essence.json" }` |
-| `decantr_context` | Read scaffold, section, page, task, and execution-pack context | `{ "action": "task", "project_path": "apps/web", "route": "/feed", "task": "improve recipe card loading" }` |
+| `decantr_context` | Read scaffold, section, page, task, shared discovery summary, and execution-pack context | `{ "action": "task", "project_path": "apps/web", "route": "/feed", "task": "improve recipe card loading" }` |
 | `decantr_graph` | Read graph snapshots, query graph nodes/edges, or traverse graph relations | `{ "action": "query", "project_path": "apps/web", "file_path": "src/app/page.tsx", "include_impact": true }` |
 | `decantr_registry` | Search and resolve hosted registry content, benchmarks, and execution packs | `{ "action": "resolve_pattern", "id": "data-table", "preset": "product" }` |
-| `decantr_verify` | Run audit, critique, findings, v2 evidence bundles, or health-loop reads | `{ "action": "evidence_bundle", "project_path": "apps/web" }` |
+| `decantr_verify` | Run audit, critique, findings, v2 evidence bundles with discovery metadata, or health-loop reads | `{ "action": "evidence_bundle", "project_path": "apps/web" }` |
 | `decantr_repair` | Return typed findings, repair plans, repair prompts, and v2 health-loop guidance | `{ "action": "health_loop", "project_path": "apps/web" }` |
 | `decantr_contract_write` | Explicit write surface for accepting drift, deferring drift to the drift log, or mutating Essence v4 | `{ "action": "update_essence", "operation": "add_feature", "payload": { "feature": "billing" } }` |
 
@@ -155,7 +167,7 @@ Registry and pack-resolution tools may call the configured Decantr API. Source u
 
 ### 3.4 Tool Surface Migration
 
-Decantr 3.4 consolidated legacy MCP tool names into the eight action-based tools above. Decantr 3.6 keeps that surface, preserves v2 task/evidence/workspace-health/health-loop contracts, and adds compatible metadata only through the consolidated tools. For example, `decantr_get_project_state` becomes `decantr_project` with `{ "action": "state" }`, `decantr_get_graph_snapshot` becomes `decantr_graph` with `{ "action": "snapshot" }`, and the legacy write tools become `decantr_contract_write` with `{ "action": "accept_drift" }` or `{ "action": "update_essence" }`.
+Decantr 3.4 consolidated legacy MCP tool names into the eight action-based tools above. Decantr 3.7 keeps that surface, preserves v2 task/evidence/workspace-health/health-loop contracts, and adds compatible discovery metadata only through the consolidated tools. For example, `decantr_get_project_state` becomes `decantr_project` with `{ "action": "state" }`, `decantr_get_graph_snapshot` becomes `decantr_graph` with `{ "action": "snapshot" }`, and the legacy write tools become `decantr_contract_write` with `{ "action": "accept_drift" }` or `{ "action": "update_essence" }`.
 
 ## How It Works
 

package/dist/bin.js

@@ -1,2 +1,2 @@
 #!/usr/bin/env node
-import "./chunk-7X4OMCHQ.js";
+import "./chunk-S3XFNHZ5.js";

package/dist/{chunk-7X4OMCHQ.js → chunk-S3XFNHZ5.js}

@@ -35,6 +35,7 @@ import {
   createEvidenceTier,
   createLoopReadiness,
   deriveVerificationDiagnostic,
+  discoverProject,
   KNOWN_VERIFICATION_DIAGNOSTICS,
   LOOP_READINESS_V2_SCHEMA_URL,
   PROJECT_HEALTH_REPORT_V2_SCHEMA_URL,
@@ -234,6 +235,65 @@ function displayProjectFile(projectRoot, path) {
   if (isAbsolute2(path)) return displayWorkspacePath(path);
   return displayWorkspacePath(join2(projectRoot, path));
 }
+function mcpDiscoverySummary(projectRoot) {
+  const discovery = discoverProject(projectRoot);
+  return {
+    schema_version: "discovery.v1",
+    project_path: discovery.workspace.projectPath,
+    workspace_scope: discovery.workspace.scope,
+    project: {
+      framework: discovery.project.framework,
+      framework_version: discovery.project.frameworkVersion,
+      package_manager: discovery.project.packageManager,
+      primary_language: discovery.project.primaryLanguage,
+      has_typescript: discovery.project.hasTypeScript,
+      has_tailwind: discovery.project.hasTailwind,
+      has_decantr: discovery.project.hasDecantr,
+      package_name: discovery.project.packageName,
+      evidence: discovery.project.evidence
+    },
+    routes: {
+      strategy: discovery.routes.strategy,
+      route_signal_count: discovery.routes.routeSignalCount,
+      taskable_route_count: discovery.routes.taskableRouteCount,
+      confidence: discovery.routes.confidence,
+      taskable_routes: discovery.routes.taskableRoutes.slice(0, 20).map((route) => ({
+        path: route.path,
+        file: route.file,
+        source: route.source,
+        confidence: route.confidence
+      })),
+      signals: discovery.routes.routeSignals.slice(0, 20).map((signal) => ({
+        path: signal.path,
+        file: signal.file,
+        kind: signal.kind,
+        taskable: signal.taskable,
+        confidence: signal.confidence
+      }))
+    },
+    components: {
+      component_count: discovery.components.componentCount,
+      page_count: discovery.components.pageCount,
+      confidence: discovery.components.confidence,
+      directories: discovery.components.directories,
+      evidence: discovery.components.evidence,
+      limitations: discovery.components.limitations
+    },
+    styling: {
+      approach: discovery.styling.approach,
+      config_file: discovery.styling.configFile,
+      css_variable_count: discovery.styling.cssVariableCount,
+      color_token_count: discovery.styling.colorTokenCount,
+      dark_mode: discovery.styling.darkMode,
+      theme_signals: discovery.styling.themeSignals
+    },
+    assistant: {
+      rule_files: discovery.assistant.ruleFiles
+    },
+    confidence: discovery.confidence,
+    limitations: discovery.limitations
+  };
+}
 function graphAvailableRoutes(snapshot) {
   return snapshot.nodes.filter((node) => node.type === "Route").map((node) => graphPayloadString(node.payload, "path") ?? node.id.replace(/^rt:/, "")).sort();
 }
@@ -2292,7 +2352,7 @@ async function handleLegacyTool(name, args) {
       if (args.components_used && Array.isArray(args.components_used) && args.page_id && typeof args.page_id === "string") {
         const pages = listEssencePages(essence);
         const page = pages.find((p) => p.id === args.page_id);
-        if (page && page.layout) {
+        if (page?.layout) {
           const expectedPatterns = /* @__PURE__ */ new Set();
           for (const item of page.layout) {
             if (typeof item === "string") {
@@ -2621,6 +2681,7 @@ async function handleLegacyTool(name, args) {
         return {
           source: "local_workspace",
           project_root: displayWorkspacePath(projectRoot),
+          discovery: mcpDiscoverySummary(projectRoot),
           essence: essence ? {
             present: true,
             version: typeof essence.version === "string" ? essence.version : null,
@@ -3651,6 +3712,7 @@ async function handleLegacyTool(name, args) {
         verifyCommand
       };
       return {
+        discovery: mcpDiscoverySummary(projectRoot),
         route: resolvedRoute,
         page_id: pageId,
         section_id: section.id,
@@ -4072,7 +4134,10 @@ async function handleLegacyTool(name, args) {
       try {
         const projectRoot = resolveMcpProjectRoot(args.project_path);
         const state = await getMcpHealthState(projectRoot);
-        return state.evidence;
+        return {
+          ...state.evidence,
+          discovery: mcpDiscoverySummary(projectRoot)
+        };
       } catch (error) {
         return { error: error.message };
       }
@@ -4111,6 +4176,7 @@ async function handleLegacyTool(name, args) {
         }
         return {
           project: state.evidence.project,
+          discovery: mcpDiscoverySummary(projectRoot),
           health: state.evidence.health,
           finding: {
             id: finding.id,
@@ -4329,7 +4395,7 @@ function describeUpdate(operation, payload) {
 }
 
 // src/index.ts
-var VERSION = "3.6.1";
+var VERSION = "3.7.0";
 var server = new Server({ name: "decantr", version: VERSION }, { capabilities: { tools: {} } });
 server.setRequestHandler(ListToolsRequestSchema, async () => {
   return { tools: TOOLS };

package/dist/index.js

@@ -1 +1 @@
-import "./chunk-7X4OMCHQ.js";
+import "./chunk-S3XFNHZ5.js";

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@decantr/mcp-server",
-  "version": "3.6.1",
+  "version": "3.7.0",
   "mcpName": "io.github.decantr-ai/mcp-server",
   "description": "MCP server for Decantr — exposes Contract context, typed graph artifacts, and verification evidence to AI coding assistants",
   "keywords": [
@@ -51,10 +51,10 @@
   },
   "dependencies": {
     "@modelcontextprotocol/sdk": "^1.29.0",
-    "@decantr/core": "3.6.0",
-    "@decantr/verifier": "3.6.1",
     "@decantr/essence-spec": "3.4.0",
-    "@decantr/registry": "3.4.0"
+    "@decantr/core": "3.6.0",
+    "@decantr/registry": "3.4.0",
+    "@decantr/verifier": "3.7.0"
   },
   "scripts": {
     "build": "tsup",