kiku-agent 1.3.1 → 1.3.3

package/README.md

@@ -52,7 +52,7 @@ kiku-agent
 ```
 
 > [!NOTE]
-> The global package deliberately excludes native binary build dependencies (Python, `node-gyp`, Visual Studio C++ build tools are not required). For a local checkout before publishing, pack and install via `npm install -g ./kiku-agent-1.3.1.tgz`.
+> The global package deliberately excludes native binary build dependencies (Python, `node-gyp`, Visual Studio C++ build tools are not required). For a local checkout before publishing, pack and install via `npm install -g ./kiku-agent-1.3.3.tgz`.
 
 ### Zero-Install One-Shot Query
 
@@ -73,9 +73,12 @@ npx kiku-agent --query "ผลงานซีรีส์ปี 2026 มีเ
    kiku-agent
    ```
 
-### Launch with Cloud Providers
+### Launch with Local Engines & Cloud Providers
 
 ```bash
+# Ollama (Local port 11434)
+npm run kiku-agent -- --provider ollama --model qwen2.5:7b
+
 # OpenAI (GPT-5.6 / GPT-4o)
 npm run kiku-agent -- --provider openai --model gpt-5.6-luna
 
@@ -85,7 +88,7 @@ npm run kiku-agent -- --provider gemini --model gemini-3.7-flash
 # Groq (Ultra-fast LPU inference)
 npm run kiku-agent -- --provider groq --model openai/gpt-oss-120b
 
-# DeepSeek (DeepSeek V4)
+# DeepSeek (DeepSeek V4 / R1)
 npm run kiku-agent -- --provider deepseek --model deepseek-v4-flash
 
 # OpenRouter (Dynamic routing)
@@ -260,38 +263,91 @@ Kiku Agent includes **60 validated tools** across 6 categories that the agent au
 
 ## Model Context Protocol (MCP)
 
-### 1. MCP Server (`kiku-mcp`)
+### 1. Dual-Transport MCP Server (`kiku-mcp` & Streamable HTTP `POST /mcp`)
 
-The package ships an official Model Context Protocol server exposing the Ju Jingyi encyclopedia as **58 tools** (55 agent tools + 3 compatibility aliases: `get_profile`, `get_timeline`, `search_gallery`), **8 resources**, and **7 prompt templates**. Live-web and Chinese tools are opt-in via `KIKU_MCP_ENABLE_WEB_TOOLS=true` and `KIKU_ENABLE_CN_WEB_TOOLS=true`:
+The package ships an official Model Context Protocol server exposing the Ju Jingyi encyclopedia as **68+ tools**, **8 resources**, and **7 prompt templates**, fully modernized to **Protocol Version `2025-11-25`** (with automatic negotiation for `2024-11-05` backwards compatibility):
 
 ```bash
-# Stdio transport (Claude Desktop, Cursor, VS Code)
+# Stdio transport (CLI command)
 kiku-mcp
 
 # Zero-install via npx:
 npx -y -p kiku-agent kiku-mcp
 ```
 
-#### Claude Desktop & Cursor Integration
+### 2. Supported AI Clients & IDEs (10 Platforms)
+
+Kiku Agent provides ready-to-use configuration templates for **10 major AI clients & IDE platforms** across both **Streamable HTTP** (recommended for zero overhead) and **Local Stdio**:
+
+| # | Client / Tool | Target Config File | Format & Spec | Category |
+|---|---|---|---|---|
+| 1 | **Claude Desktop** | `claude_desktop_config.json` | JSON (`mcpServers.kiku`) | Desktop App |
+| 2 | **Cursor** | `.cursor/mcp.json` | JSON (`mcpServers.kiku`) | AI IDE |
+| 3 | **Windsurf** | `~/.codeium/windsurf/mcp_config.json` | JSON (`mcpServers.kiku`) | AI IDE (Cascade) |
+| 4 | **VS Code & Copilot** | `.vscode/mcp.json` | JSON (`servers.kiku` with `type`) | IDE Extension |
+| 5 | **Google Antigravity (AGY)** | `.agents/mcp_config.json` | JSON (`mcpServers.kiku`) | DeepMind Agent |
+| 6 | **Cline / Roo Code** | `cline_mcp_settings.json` | JSON (`mcpServers.kiku`) | VS Code Agent |
+| 7 | **Claude Code CLI** | `~/.claude/mcp.json` | JSON or `claude mcp add kiku <url>` | CLI Agent |
+| 8 | **Gemini CLI** | `~/.gemini/mcp.json` | JSON (`mcpServers.kiku.httpUrl`) | CLI Agent |
+| 9 | **LibreChat** | `librechat.yaml` | YAML (`streamable-http` / `stdio`) | Multi-User Web UI |
+| 10 | **OpenCode / Open WebUI** | `opencode.json` | JSON (`$schema` + `mcp.kiku`) | Terminal & Web UI |
+
+> [!TIP]
+> **One-Click Copy in UI**: You can browse, preview, and copy configurations for all 10 clients directly in the **Web Chat MCP Modal (Tab 3)** or in **Settings (`/settings#mcp`)** with instant toggling between Streamable HTTP and Local Stdio.
+
+#### Quick Configuration Examples
 
-Add the following snippet to your `claude_desktop_config.json` or Cursor MCP configuration:
+**Streamable HTTP (Recommended for Claude Desktop, Cursor, Windsurf, AGY):**
+```json
+{
+  "mcpServers": {
+    "kiku": {
+      "url": "http://localhost:3002/mcp"
+    }
+  }
+}
+```
+
+**VS Code & GitHub Copilot (`.vscode/mcp.json`):**
+```json
+{
+  "servers": {
+    "kiku": {
+      "type": "http",
+      "url": "http://localhost:3002/mcp"
+    }
+  }
+}
+```
+
+**LibreChat (`librechat.yaml`):**
+```yaml
+mcpServers:
+  kiku:
+    type: streamable-http
+    url: http://localhost:3002/mcp
+```
 
+**Local Stdio (Zero-Install via NPX):**
 ```json
 {
   "mcpServers": {
     "kiku": {
       "command": "npx",
-      "args": ["-y", "-p", "kiku-agent", "kiku-mcp"]
+      "args": ["-y", "-p", "kiku-agent", "kiku-mcp"],
+      "env": {
+        "KIKU_MCP_ENABLE_WEB_TOOLS": "true"
+      }
     }
   }
 }
 ```
 
-The production deployment exposes a stateless Streamable HTTP endpoint at `https://jujingyi-kikuuu.vercel.app/mcp` (`POST /mcp`).
+The server endpoint is available locally at `http://localhost:3002/mcp` and remotely at `https://jujingyi-kikuuu.vercel.app/mcp`.
 
 > [!IMPORTANT]
 > **HTTP Security**: The remote MCP endpoint is guarded by a bearer token and CORS allow-list:
-> - Set `MCP_API_KEY` to require `Authorization: Bearer <MCP_API_KEY>` on `POST /mcp`.
+> - Set `MCP_API_KEY` to require `Authorization: Bearer <MCP_API_KEY>` or `X-API-Key` on `POST /mcp` (zero-key for local loopback).
 > - Set `MCP_CORS_ORIGINS` to a comma-separated list of allowed browser origins (empty = no CORS).
 
 ```bash
@@ -320,6 +376,7 @@ curl -X POST https://jujingyi-kikuuu.vercel.app/mcp \
 A frozen catalog QA harness in `server/agent/eval/` runs **500 fixed multilingual cases** (350 catalog & reasoning + 150 security & red-teaming across 17 categories) through a fresh real `KikuAgent` per case on local LM Studio, recording answers, tool traces, latency, and token telemetry:
 
 ```bash
+npm run eval:compat                 # 7-point enterprise model compatibility probes
 npm run eval:validate                 # Validate frozen 500-case corpus integrity
 npm run eval:quick                    # 3-case smoke run (needs LM Studio)
 npm run eval:ui                       # Launch visual Kiku Eval Studio workbench on port 3001 (/eval)
@@ -363,7 +420,7 @@ Copy `.env.example` → `.env` (parsed automatically when the CLI runs; only uns
 | `MCP_API_KEY` | Bearer token required by the HTTP MCP endpoint (`POST /mcp`) |
 | `MCP_CORS_ORIGINS` | Comma-separated allowed origins for browser MCP calls |
 | `KIKU_MCP_ENABLE_WEB_TOOLS` | Set `true` to enable web tools in the MCP server |
-| `KIKU_ENABLE_CN_WEB_TOOLS` | Set `true` to enable Chinese web search portals, Bilibili, and image search |
+| `KIKU_ENABLE_CN_WEB_TOOLS` | Enabled by default; set `false` to disable Chinese web search portals, Bilibili, and image search |
 | `KIKU_API_KEYS` | Pre-seeded REST API keys (higher rate limits) |
 | `KIKU_VIP_KEYS` | Pre-seeded Developer VIP API keys (highest rate limits) |
 

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "kiku-agent",
-  "version": "1.3.1",
+  "version": "1.3.3",
   "description": "Kiku AI terminal agent and MCP server for Ju Jingyi knowledge, media, and career data.",
   "license": "GPL-3.0",
   "type": "module",
@@ -14,17 +14,7 @@
     "./tools": "./server/agent/tools.js"
   },
   "files": [
-    "server/*.js",
-    "server/routes/**/*",
-    "server/web-chat/**/*",
-    "server/agent/*.js",
-    "server/agent/websearch/**/*",
-    "server/agent/memory/**/*",
-    "server/agent/telemetry/**/*",
-    "server/agent/ink/**/*",
-    "server/mcp/*.js",
-    "server/mcp/data/*.js",
-    "server/mcp/**/*.json",
+    "server/**/*",
     "tokens.css",
     "src/data/**/*.js",
     "src/lib/**/*.js",
@@ -74,12 +64,15 @@
     "profile:nsys": "node scripts/profileNsys.js",
     "test": "node --test server/agent/tests/*.test.js server/mcp/tests/*.test.js",
     "sync:check": "node scripts/verifyCatalogSync.js",
+    "pack:verify": "node scripts/verifyPackaging.js --thorough",
+    "prepack": "node scripts/verifyPackaging.js",
     "bump": "node bump_version.js",
     "eval": "node server/agent/eval/modules/core/cli.js",
     "eval:validate": "node server/agent/eval/modules/core/cli.js validate",
     "eval:quick": "node server/agent/eval/modules/core/cli.js run --limit 3",
     "eval:security": "node server/agent/eval/modules/core/cli.js run --suite security",
     "eval:run": "node server/agent/eval/modules/core/cli.js run",
+    "eval:compat": "node scripts/evalCompat.js",
     "eval:judge": "node server/agent/eval/modules/core/cli.js judge",
     "eval:compare": "node server/agent/eval/modules/core/cli.js compare",
     "eval:dashboard": "node server/agent/eval/modules/core/cli.js dashboard --open",