add schemas and swap browser_use_runner in docker with browser and now agent use skill from the other container

api/schemas.py

@@ -0,0 +1,15 @@
+from typing import Optional
+
+from pydantic import BaseModel, Field
+
+
+class BrowserTaskRequest(BaseModel):
+    task: str = Field(..., description="Задача для браузера")
+    timeout: int = Field(300, description="Максимальное время выполнения задачи в секундах")
+
+
+class BrowserTaskResponse(BaseModel):
+    success: bool
+    result: Optional[str] = None
+    error: Optional[str] = None
+    execution_time: float

browser_env/Dockerfile.browser

@@ -4,6 +4,8 @@ ENV DEBIAN_FRONTEND=noninteractive
 
 RUN apt-get update && apt-get install -y --no-install-recommends \
     chromium \
+    python3 \
+    python3-pip \
     xvfb \
     fluxbox \
     x11vnc \
@@ -19,9 +21,14 @@ RUN apt-get update && apt-get install -y --no-install-recommends \
 WORKDIR /src
 RUN mkdir -p /src/browser_data
 
+RUN python3 -m pip install --no-cache-dir --break-system-packages \
+    "browser-use>=0.12.5" \
+    "langchain-openai>=0.3.0"
+
 COPY entrypoint.sh /entrypoint.sh
+COPY browser_use_runner.py /src/browser_use_runner.py
 RUN chmod +x /entrypoint.sh
 
-EXPOSE 6080 9222
+EXPOSE 6080 9222 8787
 
 ENTRYPOINT ["/entrypoint.sh"]

browser_env/browser_use_runner.py

@@ -0,0 +1,102 @@
+import asyncio
+import json
+import os
+from http.server import BaseHTTPRequestHandler, ThreadingHTTPServer
+from urllib import error, request
+
+from browser_use import Agent, Browser, ChatOpenAI
+
+
+def _json_response(handler, status_code, payload):
+    data = json.dumps(payload, ensure_ascii=False).encode("utf-8")
+    handler.send_response(status_code)
+    handler.send_header("Content-Type", "application/json; charset=utf-8")
+    handler.send_header("Content-Length", str(len(data)))
+    handler.end_headers()
+    handler.wfile.write(data)
+
+
+async def run_browser_task(task):
+    cdp_url = os.getenv("BROWSER_CDP_URL", "http://127.0.0.1:9222")
+    browser_view_url = os.getenv("BROWSER_VIEW_URL", "")
+
+    browser = Browser(cdp_url=cdp_url)
+
+    llm = ChatOpenAI(
+        model=os.getenv("MODEL_DEFAULT", "qwen3.5-122b"),
+        api_key=os.getenv("OPENAI_API_KEY"),
+        base_url=os.getenv("OPENAI_BASE_URL"),
+        temperature=0.0,
+    )
+
+    agent = Agent(task=task, llm=llm, browser=browser, use_vision=False)
+
+    try:
+        history = await agent.run()
+        return {
+            "success": True,
+            "result": history.final_result(),
+            "browser_view": browser_view_url,
+        }
+    except Exception as err:
+        return {"success": False, "error": f"Browser automation failed: {err}"}
+    finally:
+        try:
+            await browser.close()
+        except Exception:
+            pass
+
+
+class BrowserUseRPCHandler(BaseHTTPRequestHandler):
+    def do_GET(self):
+        if self.path != "/health":
+            _json_response(self, 404, {"success": False, "error": "Not found"})
+            return
+
+        try:
+            debug_url = os.getenv("BROWSER_HEALTH_URL", "http://127.0.0.1:9222/json/version")
+            with request.urlopen(debug_url, timeout=2):
+                pass
+            _json_response(self, 200, {"success": True})
+        except Exception as err:
+            _json_response(self, 503, {"success": False, "error": f"Browser is not ready: {err}"})
+
+    def do_POST(self):
+        if self.path != "/run":
+            _json_response(self, 404, {"success": False, "error": "Not found"})
+            return
+
+        try:
+            content_length = int(self.headers.get("Content-Length", "0"))
+            raw = self.rfile.read(content_length)
+            payload = json.loads(raw.decode("utf-8") if raw else "{}")
+            task = payload.get("task", "")
+            if not isinstance(task, str) or not task.strip():
+                _json_response(self, 400, {"success": False, "error": "Field 'task' is required"})
+                return
+
+            result = asyncio.run(run_browser_task(task.strip()))
+            code = 200 if result.get("success") else 500
+            _json_response(self, code, result)
+        except json.JSONDecodeError:
+            _json_response(self, 400, {"success": False, "error": "Invalid JSON payload"})
+        except error.URLError as err:
+            _json_response(self, 503, {"success": False, "error": f"Transport error: {err}"})
+        except Exception as err:
+            _json_response(self, 500, {"success": False, "error": f"Internal error: {err}"})
+
+    def log_message(self, format_str, *args):
+        return
+
+
+def main():
+    host = os.getenv("BROWSER_USE_RPC_HOST", "0.0.0.0")
+    port = int(os.getenv("BROWSER_USE_RPC_PORT", "8787"))
+    server = ThreadingHTTPServer((host, port), BrowserUseRPCHandler)
+    print(f"browser-use RPC listening on {host}:{port}")
+    server.serve_forever()
+
+
+if __name__ == "__main__":
+    main()
+

browser_env/entrypoint.sh

@@ -9,6 +9,7 @@ VNC_PORT="${VNC_PORT:-5900}"
 NOVNC_PORT="${NOVNC_PORT:-6080}"
 CHROME_LOCAL_DEBUG_PORT="${CHROME_LOCAL_DEBUG_PORT:-9223}"
 CHROME_PUBLIC_DEBUG_PORT="${CHROME_PUBLIC_DEBUG_PORT:-9222}"
+BROWSER_USE_RPC_PORT="${BROWSER_USE_RPC_PORT:-8787}"
 CHROME_PROFILE_DIR="${CHROME_PROFILE_DIR:-/src/browser_data}"
 
 MAX_RESTARTS="${MAX_RESTARTS:-10}"
@@ -119,6 +120,7 @@ start_bg fluxbox
 start_bg x11vnc -display "$DISPLAY" -rfbport "$VNC_PORT" -nopw -listen 0.0.0.0 -xkb -forever -shared
 start_bg websockify --web=/usr/share/novnc/ "$NOVNC_PORT" "localhost:${VNC_PORT}"
 start_bg socat "TCP-LISTEN:${CHROME_PUBLIC_DEBUG_PORT},fork,reuseaddr" "TCP:127.0.0.1:${CHROME_LOCAL_DEBUG_PORT}"
+start_bg python3 -u /src/browser_use_runner.py
 
 if ! wait_for_port 127.0.0.1 "$VNC_PORT" 20; then
   log "fatal: x11vnc did not open port ${VNC_PORT}"
@@ -128,8 +130,12 @@ if ! wait_for_port 127.0.0.1 "$NOVNC_PORT" 20; then
   log "fatal: websockify did not open port ${NOVNC_PORT}"
   exit 1
 fi
+if ! wait_for_port 127.0.0.1 "$BROWSER_USE_RPC_PORT" 20; then
+  log "fatal: browser-use RPC did not open port ${BROWSER_USE_RPC_PORT}"
+  exit 1
+fi
 
-log "browser infrastructure is ready (noVNC:${NOVNC_PORT}, DevTools proxy:${CHROME_PUBLIC_DEBUG_PORT})"
+log "browser infrastructure is ready (noVNC:${NOVNC_PORT}, DevTools proxy:${CHROME_PUBLIC_DEBUG_PORT}, browser-use RPC:${BROWSER_USE_RPC_PORT})"
 
 while true; do
   for pid in "${PIDS[@]}"; do

docker-compose.yml

@@ -8,6 +8,7 @@ services:
       - .env
     environment:
       - BROWSER_URL=http://browser:9222
+      - BROWSER_USE_RPC_URL=http://browser:8787/run
       - HERMES_HOME=/app/hermes_data
     volumes:
       - ./hermes_code:/app/hermes_code:ro
@@ -40,6 +41,8 @@ services:
       context: ./browser_env
       dockerfile: Dockerfile.browser
     container_name: hermes-browser
+    env_file:
+      - .env
     ports:
       - "6080:6080"
       - "9222:9222"
@@ -52,7 +55,7 @@ services:
       - browser_profiles:/src/browser_data
     restart: always
     healthcheck:
-      test: ["CMD-SHELL", "curl -fsS http://127.0.0.1:9222/json/version >/dev/null || exit 1"]
+      test: ["CMD-SHELL", "curl -fsS http://127.0.0.1:9222/json/version >/dev/null && curl -fsS http://127.0.0.1:8787/health >/dev/null || exit 1"]
       interval: 10s
       timeout: 3s
       retries: 12

hermes_code/package.json

@@ -4,7 +4,7 @@
   "description": "An AI agent with advanced tool-calling capabilities, featuring a flexible toolsets system for organizing and managing tools.",
   "private": true,
   "scripts": {
-    "postinstall": "echo '✅ Browser tools ready. Run: python run_agent.py --help'"
+    "postinstall": "echo '✅ Hermes Node hooks ready. Run: python run_agent.py --help'"
   },
   "repository": {
     "type": "git",
@@ -15,9 +15,7 @@
     "url": "https://github.com/NousResearch/Hermes-Agent/issues"
   },
   "homepage": "https://github.com/NousResearch/Hermes-Agent#readme",
-  "dependencies": {
-    "agent-browser": "^0.13.0"
-  },
+  "dependencies": {},
   "engines": {
     "node": ">=18.0.0"
   }

hermes_code/pyproject.toml

@@ -34,10 +34,6 @@ dependencies = [
   "faster-whisper>=1.0.0,<2",
   # Skills Hub (GitHub App JWT auth — optional, only needed for bot identity)
   "PyJWT[crypto]>=2.10.1,<3",
-  "browser-use>=0.12.5",
-  "playwright>=1.49.0",
-  "playwright-stealth>=1.0.6",
-  "langchain-openai>=1.1.12",
 ]
 
 [project.optional-dependencies]

hermes_code/requirements.txt

@@ -19,9 +19,6 @@ PyJWT[crypto]
 # Web tools
 firecrawl-py
 parallel-web>=0.4.2
-browser-use>=0.12.5
-playwright
-playwright-stealth
 
 # Image generation
 fal-client
@@ -36,5 +33,4 @@ croniter
 python-telegram-bot>=20.0
 discord.py>=2.0
 aiohttp>=3.9.0
-langchain-openai>=1.1.12,
 

hermes_code/scripts/install.sh

@@ -874,41 +874,14 @@ install_node_deps() {
     fi
 
     if [ -f "$INSTALL_DIR/package.json" ]; then
-        log_info "Installing Node.js dependencies (browser tools)..."
+        log_info "Installing Node.js dependencies..."
         cd "$INSTALL_DIR"
         npm install --silent 2>/dev/null || {
-            log_warn "npm install failed (browser tools may not work)"
+            log_warn "npm install failed (some Node.js helpers may not work)"
         }
         log_success "Node.js dependencies installed"
 
-        # Install Playwright browser + system dependencies.
-        # Playwright's install-deps only supports apt/dnf/zypper natively.
-        # For Arch/Manjaro we install the system libs via pacman first.
-        log_info "Installing browser engine (Playwright Chromium)..."
-        case "$DISTRO" in
-            arch|manjaro)
-                if command -v pacman &> /dev/null; then
-                    log_info "Arch/Manjaro detected — installing Chromium system dependencies via pacman..."
-                    if command -v sudo &> /dev/null && sudo -n true 2>/dev/null; then
-                        sudo NEEDRESTART_MODE=a pacman -S --noconfirm --needed \
-                            nss atk at-spi2-core cups libdrm libxkbcommon mesa pango cairo alsa-lib >/dev/null 2>&1 || true
-                    elif [ "$(id -u)" -eq 0 ]; then
-                        pacman -S --noconfirm --needed \
-                            nss atk at-spi2-core cups libdrm libxkbcommon mesa pango cairo alsa-lib >/dev/null 2>&1 || true
-                    else
-                        log_warn "Cannot install browser deps without sudo. Run manually:"
-                        log_warn "  sudo pacman -S nss atk at-spi2-core cups libdrm libxkbcommon mesa pango cairo alsa-lib"
-                    fi
-                fi
-                cd "$INSTALL_DIR" && npx playwright install chromium 2>/dev/null || true
-                ;;
-            *)
-                log_info "Playwright may request sudo to install browser system dependencies (shared libraries)."
-                log_info "This is standard Playwright setup — Hermes itself does not require root access."
-                cd "$INSTALL_DIR" && npx playwright install --with-deps chromium 2>/dev/null || true
-                ;;
-        esac
-        log_success "Browser engine installed"
+        log_info "Skipping local Playwright/Chromium bootstrap (browser automation runs in browser container)"
     fi
 
     # Install WhatsApp bridge dependencies

hermes_code/tools/browser_use_tool.py

@@ -1,61 +1,40 @@
 import json
 import os
-import asyncio
-import socket
-from browser_use import Agent, Browser, ChatOpenAI
+from urllib import error, request
 from tools.registry import registry
 
 
-async def run_browser_task(task):
-    browser_host = "browser"
-    browser_port = 9222
-    BROWSER_VIEW_URL = os.getenv("BROWSER_VIEW_URL", "")
+def run_browser_task(task):
+    if not task or not str(task).strip():
+        return json.dumps({"success": False, "error": "Task is required"}, ensure_ascii=False)
+
+    rpc_url = os.getenv("BROWSER_USE_RPC_URL", "http://browser:8787/run")
+    timeout_sec = int(os.getenv("BROWSER_USE_RPC_TIMEOUT", "900"))
+    payload = json.dumps({"task": task}).encode("utf-8")
+    req = request.Request(rpc_url, data=payload, headers={"Content-Type": "application/json"}, method="POST")
 
     try:
-        browser_ip = socket.gethostbyname(browser_host)
-        cdp_url = f"http://{browser_ip}:{browser_port}"
-    except Exception:
-        cdp_url = f"http://{browser_host}:{browser_port}"
-
-    browser = Browser(cdp_url=cdp_url)
-
-    llm = ChatOpenAI(
-        model=os.getenv("MODEL_DEFAULT", "qwen3.5-122b"),
-        api_key=os.getenv("OPENAI_API_KEY"),
-        base_url=os.getenv("OPENAI_BASE_URL"),
-        temperature=0.0,
-    )
-
-    agent = Agent(
-        task=task,
-        llm=llm,
-        browser=browser,
-        use_vision=False
-    )
-
-    try:
-        history = await agent.run()
-        final_result = history.final_result()
-
-        response = {
-            "success": True,
-            "result": final_result,
-            "browser_view": BROWSER_VIEW_URL
-        }
-        return json.dumps(response, ensure_ascii=False)
-        
-    except Exception as e:
-        return json.dumps({
-            "success": False, 
-            "error": f"Browser automation failed: {str(e)}"
-        }, ensure_ascii=False)
-        
-    finally:
-        if browser:
-            try:
-                await browser.close()
-            except Exception:
-                pass
+        with request.urlopen(req, timeout=timeout_sec) as resp:
+            body = resp.read().decode("utf-8")
+            return body
+    except error.HTTPError as http_err:
+        body = http_err.read().decode("utf-8", errors="replace")
+        return json.dumps(
+            {
+                "success": False,
+                "error": f"browser-use RPC returned HTTP {http_err.code}",
+                "details": body,
+            },
+            ensure_ascii=False,
+        )
+    except Exception as err:
+        return json.dumps(
+            {
+                "success": False,
+                "error": f"browser-use RPC request failed: {err}",
+            },
+            ensure_ascii=False,
+        )
 
 
 registry.register(
@@ -81,6 +60,6 @@ registry.register(
         }
     },
  
-    handler=lambda args, **kw: asyncio.run(run_browser_task(args.get("task"))),
+    handler=lambda args, **kw: run_browser_task(args.get("task")),
     emoji="🌐",
 )