robot-resources 1.7.6 → 1.7.8

package/README.md

@@ -2,103 +2,88 @@
 
 > Tools for AI agents. Humans have HR. Agents have RR.
 
-Robot Resources builds tools for AI agents and any software that makes LLM API calls — chatbots, RAG pipelines, AI-powered apps, internal tools. Two products today: **Router** (intelligent model routing, 60-90% cost savings) and **Scraper** (token compression for web content, median 91% token reduction). Both run locally, both free.
+Two products for any software that makes LLM API calls — chatbots, RAG pipelines, AI-powered apps, agent runtimes. Both run locally, both free.
 
-## Quick Start
+**Router** — Intelligent LLM routing proxy. Classifies each prompt by task type, routes to the cheapest model that qualifies. 60-90% cost savings with zero quality loss.
+
+**Scraper** — Token compression for web content. Fetches any URL, strips noise, returns clean markdown. Median 91% token reduction.
+
+## Install
 
 ```bash
 npx robot-resources
 ```
 
-One command: authenticates via GitHub OAuth, installs Router as an always-on service, registers Scraper as the default web compression tool, and configures your agent automatically.
+One command: installs Router as an always-on system service, registers Scraper as an MCP tool, configures your agent automatically.
 
-## Products
+## What It Does
 
-### Router
+1. **Installs Router** — Python venv + system service on localhost:3838
+2. **Registers Scraper** — MCP tool `scraper_compress_url(url)` in your agent
+3. **Configures OpenClaw** — Plugin auto-installed if OpenClaw is detected
+4. **Provisions API key** — For telemetry and dashboard access
+5. **Health check** — Verifies everything is running after install
 
-Intelligent LLM routing proxy. Classifies each prompt by task type, filters by model capability, routes to the cheapest model that qualifies.
+## Router
 
-- Hybrid classification: keyword detection (~5ms) + LLM fallback for ambiguous prompts (~200ms)
-- Dynamic thresholds: simple tasks open cheap models (0.60), complex tasks require top models (0.85)
-- 11 models across OpenAI, Anthropic, and Google — routes within your available providers
-- 60-90% cost savings with zero quality loss
-- OpenAI-compatible API on localhost:3838 — change your base_url and you're done
-- OpenClaw compatible — plugin auto-installs, works with API keys and subscriptions
+Transparent proxy on localhost:3838. Your LLM calls pass through, Router selects the cheapest capable model:
 
-```bash
-npx @robot-resources/router
-```
+- Hybrid classification: keyword detection (~5ms) + LLM fallback (~200ms)
+- Dynamic thresholds: simple tasks → cheap models, complex tasks → top models
+- Models across OpenAI, Anthropic, Google — routes within your available providers
+- Your API keys pass through via headers — never stored by Router
+- OpenAI-compatible API — change `base_url` to `http://localhost:3838`
 
-### Scraper
+## Scraper
 
-Token compression for web content. Fetches any URL, strips noise, returns clean markdown with token count.
+Available as MCP tool after install:
 
 - Mozilla Readability extraction (0.97 F1 accuracy)
-- Content-aware token estimation (±15% of actual BPE)
+- Content-aware token estimation
 - 3-tier fetch: fast, stealth (TLS fingerprint), render (headless browser)
 - Multi-page BFS crawl with robots.txt compliance
 - Median 91% token reduction per page
 
-Installed automatically via `npx robot-resources`. Available as MCP tool `scraper_compress_url(url)` in your agent.
+## Agent Compatibility
 
-### Dashboard
+| Agent | Integration | Status |
+|-------|-------------|--------|
+| **OpenClaw** | Plugin (auto-install) | Verified |
+| **Claude Code** | MCP server | Verified |
+| **Any OpenAI client** | HTTP proxy (localhost:3838) | Compatible |
+| **Cursor** | MCP server | Compatible |
+| **Windsurf** | MCP server | Compatible |
 
-Usage dashboard with real-time telemetry. Auth via GitHub OAuth, KPI panels, routing stats, cost savings tracking. Mobile-responsive.
-
-### MCP Servers
-
-Both products include MCP servers for AI agent integration:
+## MCP Servers
 
 ```bash
 npx -y @robot-resources/router-mcp    # Router stats + config
 npx -y @robot-resources/scraper-mcp   # Scraper compression
 ```
 
-## Pricing
-
-Free. Unlimited. No tiers. Both tools run locally — you pay your AI providers directly. No markup, no rate limits, no quotas. Your API keys never leave your machine.
+## Dashboard
 
-## Monorepo Structure
+Usage dashboard at https://robotresources.ai/dashboard — real-time telemetry, cost savings, routing stats. Auth via GitHub OAuth.
 
-```
-robot-resources/
-├── router/         # LLM routing proxy (Python + TypeScript MCP)
-├── scraper/        # Token compression (TypeScript)
-├── web/            # Landing page + dashboard (React/Vite)
-├── platform/       # Backend API — auth, telemetry (Hono/Cloudflare Workers)
-├── packages/       # Unified CLI installer (npx robot-resources)
-├── skills/         # Agent skills (ClawHub)
-├── _orchestrator/  # Business coordination
-└── _brand/         # Brand assets and design system
-```
-
-## Development
-
-Each product has its own setup:
+## Pricing
 
-- [Router](./router/README.md)
-- [Scraper](./scraper/packages/scraper/README.md)
-- [Platform](./platform/README.md)
-- [CLI](./packages/cli/README.md)
+Free. Unlimited. No tiers. Your API keys never leave your machine.
 
-## CI/CD
+## Telemetry
 
-Unified pipeline on push/PR to main:
+Anonymous usage telemetry (model selection, cost savings) to improve the product. No personal data, no request content, no API keys.
 
-- **Router**: ruff, mypy, pytest
-- **Scraper**: tsc, vitest
-- **Web**: tsc, eslint
-- **Publish**: 7 npm packages via OIDC (`router`, `router-mcp`, `scraper`, `scraper-mcp`, `scraper-tracking`, `cli-core`, `cli`)
+Opt out: `export RR_TELEMETRY=off`
 
 ## Links
 
 - Website: https://robotresources.ai
-- GitHub: https://github.com/robot-resources
-- npm: `npx robot-resources`
-- Twitter/X: https://x.com/robotresources
+- Dashboard: https://robotresources.ai/dashboard
+- Agent docs: https://robotresources.ai/llms.txt
+- npm: https://www.npmjs.com/package/robot-resources
+- GitHub: https://github.com/robot-resources/packages
 - Discord: https://robotresources.ai/discord
 - Contact: agent@robotresources.ai
-- Agent docs: https://robotresources.ai/llms.txt
 
 ## License
 

package/lib/tool-config.js

@@ -1,3 +1,4 @@
+import { spawn } from 'node:child_process';
 import { createRequire } from 'node:module';
 import { readFileSync, writeFileSync, copyFileSync, mkdirSync, existsSync } from 'node:fs';
 import { homedir } from 'node:os';
@@ -220,5 +221,49 @@ export function configureToolRouting() {
   return results;
 }
 
+/**
+ * Run a command with a heartbeat to keep agent sessions alive.
+ * OC kills processes after 5s of no output (noOutputTimeoutMs = 5000).
+ * Prints immediately, then every 4s (safely under the 5s threshold).
+ */
+function spawnWithHeartbeat(cmd, args, { label, timeout = 30_000 } = {}) {
+  return new Promise((resolve, reject) => {
+    const proc = spawn(cmd, args, {
+      stdio: ['ignore', 'pipe', 'pipe'],
+      timeout,
+    });
+
+    process.stdout.write(`  ${label}...\n`);
+    let seconds = 0;
+    const heartbeat = setInterval(() => {
+      seconds += 4;
+      process.stdout.write(`  ${label}... ${seconds}s\n`);
+    }, 4000);
+
+    proc.on('close', (code) => {
+      clearInterval(heartbeat);
+      if (code === 0) resolve();
+      else reject(new Error(`${cmd} ${args.join(' ')} exited with code ${code}`));
+    });
+
+    proc.on('error', (err) => {
+      clearInterval(heartbeat);
+      reject(err);
+    });
+  });
+}
+
+/**
+ * Restart the OpenClaw gateway so it picks up new plugin + config.
+ * Uses heartbeat to keep OC sessions alive during the restart.
+ * Telegram survives this restart — tested end-to-end (PR #89).
+ */
+async function restartOpenClawGateway() {
+  await spawnWithHeartbeat('openclaw', ['gateway', 'restart'], {
+    label: 'Restarting gateway',
+    timeout: 15_000,
+  });
+}
+
 // Exported for testing and direct use
-export { stripJson5, configureOpenClaw, registerScraperMcp };
+export { stripJson5, configureOpenClaw, registerScraperMcp, restartOpenClawGateway };

package/lib/wizard.js

@@ -2,11 +2,11 @@ import { readFileSync, writeFileSync, mkdirSync } from 'node:fs';
 import { join } from 'node:path';
 import { homedir, hostname } from 'node:os';
 import { readConfig, writeConfig } from '@robot-resources/cli-core/config.mjs';
-import { findPython, isPortAvailable, isHeadless } from './detect.js';
+import { findPython, isPortAvailable, isHeadless, isOpenClawInstalled } from './detect.js';
 import { getOrCreateMachineId } from './machine-id.js';
 import { setupRouter, isRouterInstalled, getVenvPythonPath } from './python-bridge.js';
 import { installService, isServiceRunning, isServiceInstalled } from './service.js';
-import { configureToolRouting, registerScraperMcp } from './tool-config.js';
+import { configureToolRouting, registerScraperMcp, restartOpenClawGateway } from './tool-config.js';
 import { checkHealth } from './health-report.js';
 import { header, step, success, warn, error, info, blank, summary } from './ui.js';
 /**
@@ -382,11 +382,19 @@ export async function runWizard({ nonInteractive = false } = {}) {
     }
   }
 
-  // ── Gateway restart: NOT needed ─────────────────────────────────────────
+  // ── Best-effort: Single gateway restart (MUST be last) ─────────────────
   //
-  // OC has a built-in config watcher. When openclaw.json changes (plugin
-  // entry, scraper MCP, plugins.allow), the gateway detects the change,
-  // waits for active operations to complete, then restarts gracefully.
-  // No manual restart needed — the wizard just writes config and OC
-  // handles the rest.
+  // Everything is already installed, configured, and reported above.
+  // The restart loads the plugin + scraper MCP into the running gateway.
+  // Telegram survives this restart. If the session dies here, the agent
+  // picks up on the next message with all tools loaded.
+
+  if (isOpenClawInstalled() && (results.tools?.some(r => r.action === 'installed') || scraperRegistered)) {
+    try {
+      await restartOpenClawGateway();
+      success('OpenClaw gateway restarted');
+    } catch {
+      // Best-effort — gateway picks up changes on next restart
+    }
+  }
 }

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "robot-resources",
-  "version": "1.7.6",
+  "version": "1.7.8",
   "description": "Robot Resources — AI agent tools. One command to install everything.",
   "type": "module",
   "bin": {
@@ -19,7 +19,7 @@
   "dependencies": {
     "@robot-resources/cli-core": "*",
     "@robot-resources/openclaw-plugin": "*",
-    "@robot-resources/scraper": "^0.3.0"
+    "@robot-resources/scraper": "^0.3.1"
   },
   "devDependencies": {
     "vitest": "^1.2.0"
@@ -41,7 +41,7 @@
   "author": "Robot Resources Team",
   "repository": {
     "type": "git",
-    "url": "git+https://github.com/robot-resources/robot-resources.git",
+    "url": "git+https://github.com/robot-resources/packages.git",
     "directory": "packages/cli"
   },
   "homepage": "https://robotresources.ai",