taru-mcp 0.1.4 → 0.1.5

package/README.md

@@ -6,18 +6,86 @@ Zero dependencies. Pure Node.js. Works with any MCP client.
 
 ## Quick Setup
 
+### macOS / Linux
+
 ```bash
 curl -fsSL https://raw.githubusercontent.com/arupa-inc/taru-mcp/main/setup.sh | bash
 ```
 
+### Windows (PowerShell)
+
+```powershell
+irm https://raw.githubusercontent.com/arupa-inc/taru-mcp/main/setup.ps1 | iex
+```
+
 The script will ask you:
 1. **Project folder name** — creates the workspace directory
 2. **AI client** — Claude Code, Codex, or both
-3. **API token** — optional, can add later
+3. **Workspace token** — taru workspace token (`xxv_...`). Can be entered later if you don't have it yet.
 
 It handles `npm init`, `taru-mcp` installation, agent file (`CLAUDE.md` / `AGENTS.md`) setup, and MCP registration automatically.
 
-Get your API token from the taru web console: **Settings > API Key**.
+Get your workspace token from the taru web console: **Settings > API Key**.
+
+## Manual Setup
+
+If the setup script doesn't work, follow these steps:
+
+### 1. Install
+
+```bash
+mkdir my-project && cd my-project
+npm init -y
+npm install taru-mcp
+```
+
+### 2. Copy agent instructions
+
+```bash
+# For Claude Code
+cp node_modules/taru-mcp/samples/CLAUDE.md ./CLAUDE.md
+
+# For Codex
+cp node_modules/taru-mcp/samples/AGENTS.md ./AGENTS.md
+```
+
+### 3. Register MCP server
+
+```bash
+# Claude Code
+claude mcp add taru -- npx taru-mcp --url https://taru-api.arupa.io --token xxv_your_token
+
+# Codex
+codex mcp add taru -- npx taru-mcp --url https://taru-api.arupa.io --token xxv_your_token
+```
+
+### 4. Or configure MCP manually
+
+If `claude mcp add` doesn't work, create `.mcp.json` in your project root:
+
+```json
+{
+  "mcpServers": {
+    "taru": {
+      "command": "npx",
+      "args": ["taru-mcp", "--url", "https://taru-api.arupa.io", "--token", "xxv_your_token"]
+    }
+  }
+}
+```
+
+For Claude Code global config (`~/.claude.json`):
+
+```json
+{
+  "mcpServers": {
+    "taru": {
+      "command": "npx",
+      "args": ["taru-mcp", "--url", "https://taru-api.arupa.io", "--token", "xxv_your_token"]
+    }
+  }
+}
+```
 
 ## Tools
 
@@ -30,8 +98,6 @@ Once connected, these MCP tools are available to the AI:
 | `store_document` | Store a document or opinion with auto-conflict detection |
 | `list_documents` | List all documents in the workspace |
 | `list_conflicts` | View pending knowledge conflicts |
-| `web_search` | Search the web via DuckDuckGo |
-| `web_fetch` | Fetch and extract text from a URL |
 | `rebalance` | Merge similar keywords, clean up orphan nodes |
 
 ## How it works

package/bin/taru-mcp.mjs

@@ -39,7 +39,9 @@ Examples:
 }
 
 url = url.replace(/\/+$/, "");
-const endpoint = `${url}/mcp/${workspaceId}`;
+// If workspace ID was explicitly provided, use the scoped endpoint; otherwise use token-based auto endpoint
+const hasExplicitWorkspace = args.some(a => a === "--workspace-id" || a === "-w") || env.TARU_WORKSPACE_ID;
+const endpoint = hasExplicitWorkspace ? `${url}/mcp/${workspaceId}` : `${url}/mcp`;
 const isHttps = endpoint.startsWith("https://");
 
 stderr.write(`[taru-mcp] endpoint: ${endpoint}\n`);

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "taru-mcp",
-  "version": "0.1.4",
+  "version": "0.1.5",
   "description": "MCP server for taru knowledge graph — connect Claude Code or Codex to your team's shared brain",
   "bin": {
     "taru-mcp": "./bin/taru-mcp.mjs"

package/samples/AGENTS.md

@@ -9,8 +9,6 @@ You have access to the taru MCP server — a research knowledge graph with tempo
 - store_document: Store extracted knowledge (handles embedding + graph)
 - list_documents: List all documents with status and confidence
 - list_conflicts: View pending knowledge conflicts
-- web_search: Search the web (DuckDuckGo)
-- web_fetch: Fetch text content from a URL
 - rebalance: Clean up the knowledge graph (merge keywords, remove orphans)
 
 ### Rules

package/samples/CLAUDE.md

@@ -9,8 +9,6 @@ You have access to the taru MCP server — a research knowledge graph with tempo
 - store_document: Store extracted knowledge (handles embedding + graph)
 - list_documents: List all documents with status and confidence
 - list_conflicts: View pending knowledge conflicts
-- web_search: Search the web (DuckDuckGo)
-- web_fetch: Fetch text content from a URL
 - rebalance: Clean up the knowledge graph (merge keywords, remove orphans)
 
 ### Rules

package/setup.sh

@@ -12,7 +12,7 @@ echo "  🌳 taru MCP Setup"
 echo ""
 
 # 1. Ask project folder name
-read -rp "Project folder name: " FOLDER
+read -rp "Project folder name: " FOLDER </dev/tty
 if [ -z "$FOLDER" ]; then
   echo "Error: folder name is required"
   exit 1
@@ -33,7 +33,7 @@ echo "  1) Claude Code"
 echo "  2) Codex"
 echo "  3) Not sure (copies both CLAUDE.md and AGENTS.md)"
 echo ""
-read -rp "Choose [1/2/3]: " CLIENT_CHOICE
+read -rp "Choose [1/2/3]: " CLIENT_CHOICE </dev/tty
 
 case "$CLIENT_CHOICE" in
   1) CLIENT="claude" ;;
@@ -43,7 +43,7 @@ esac
 
 # 3. Ask token (skippable)
 echo ""
-read -rp "API token (press Enter to skip): " TOKEN
+read -rp "Workspace token (xxv_..., press Enter to add later): " TOKEN </dev/tty
 
 # Check npm
 if ! command -v npm &>/dev/null; then
@@ -114,7 +114,7 @@ echo ""
 echo "  Project: $(pwd)"
 [ "$CLIENT" = "claude" ] || [ "$CLIENT" = "both" ] && echo "  Agent file: ./CLAUDE.md"
 [ "$CLIENT" = "codex" ] || [ "$CLIENT" = "both" ] && echo "  Agent file: ./AGENTS.md"
-[ -n "$TOKEN" ] && echo "  Token: configured" || echo "  Token: skipped (add later via settings)"
+[ -n "$TOKEN" ] && echo "  Token: configured" || echo "  Token: not set yet (add via settings)"
 echo ""
 echo "  Open this folder in your AI client to start growing the tree."
 echo ""