agnt 0.7.6__tar.gz

agnt-0.7.6/.gitignore

@@ -0,0 +1,51 @@
+# Binaries (root level only, not cmd/devtool-mcp/)
+/devtool-mcp
+*.exe
+*.dll
+*.so
+*.dylib
+
+# Test binary, built with `go test -c`
+*.test
+
+# Output of the go coverage tool
+*.out
+coverage.html
+
+# Go workspace file
+go.work
+
+# IDE
+.idea/
+.vscode/
+*.swp
+*.swo
+
+# OS
+.DS_Store
+Thumbs.db
+
+# Documentation build
+docs-site/node_modules/
+docs-site/build/
+docs-site/.docusaurus/
+
+# Environment
+.env
+.env.local
+
+# npm package
+node_modules/
+bin/devtool-mcp
+bin/devtool-mcp.exe
+dist/
+*.tgz
+
+# Python package
+python/dist/
+python/build/
+python/*.egg-info/
+__pycache__/
+*.pyc
+.venv/
+.mcpregistry_*

agnt-0.7.6/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2024 Standard Beagle
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

agnt-0.7.6/PKG-INFO

@@ -0,0 +1,187 @@
+Metadata-Version: 2.4
+Name: agnt
+Version: 0.7.6
+Summary: MCP server for AI coding agents - process management, reverse proxy with traffic logging, browser instrumentation, and sketch mode
+Project-URL: Homepage, https://standardbeagle.github.io/agnt/
+Project-URL: Documentation, https://standardbeagle.github.io/agnt/
+Project-URL: Repository, https://github.com/standardbeagle/agnt
+Project-URL: Issues, https://github.com/standardbeagle/agnt/issues
+Project-URL: Changelog, https://github.com/standardbeagle/agnt/blob/main/CHANGELOG.md
+Author-email: Standard Beagle <dev@standardbeagle.com>
+License-Expression: MIT
+License-File: LICENSE
+Keywords: agent,ai,claude-code,debugging,development,devtools,diagnostics,frontend,mcp,model-context-protocol,process-management,proxy,sketch,wireframe
+Classifier: Development Status :: 4 - Beta
+Classifier: Environment :: Console
+Classifier: Intended Audience :: Developers
+Classifier: License :: OSI Approved :: MIT License
+Classifier: Operating System :: MacOS
+Classifier: Operating System :: Microsoft :: Windows
+Classifier: Operating System :: POSIX :: Linux
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Topic :: Internet :: Proxy Servers
+Classifier: Topic :: Software Development :: Debuggers
+Classifier: Topic :: Software Development :: Testing
+Requires-Python: >=3.10
+Requires-Dist: httpx>=0.25.0
+Description-Content-Type: text/markdown
+
+# agnt
+
+**Give your AI coding agent browser superpowers.**
+
+agnt is a new kind of tool designed for the age of AI-assisted development. It acts as a bridge between your AI coding agent and the browser, extending what's possible during vibe coding sessions.
+
+## What Does It Do?
+
+When you're in the flow with Claude Code, Cursor, or other AI tools, agnt lets your agent:
+
+- **See what you see** - Screenshots, DOM inspection, visual debugging
+- **Hear from you directly** - Send messages from browser to agent
+- **Sketch ideas together** - Draw wireframes directly on your UI
+- **Debug in real-time** - Capture errors, network traffic, performance
+- **Extend its thinking window** - Structured data uses fewer tokens than descriptions
+
+## Installation
+
+```bash
+pip install agnt
+# or
+uv pip install agnt
+# or
+pipx install agnt
+```
+
+## Quick Start
+
+### As MCP Server (Claude Code, Cursor, etc.)
+
+Add to your MCP configuration:
+
+```json
+{
+  "mcpServers": {
+    "agnt": {
+      "command": "agnt",
+      "args": ["serve"]
+    }
+  }
+}
+```
+
+Or with uvx:
+
+```json
+{
+  "mcpServers": {
+    "agnt": {
+      "command": "uvx",
+      "args": ["agnt", "serve"]
+    }
+  }
+}
+```
+
+### As PTY Wrapper
+
+Wrap your AI coding tool with overlay features:
+
+```bash
+agnt run claude --dangerously-skip-permissions
+agnt run cursor
+agnt run aider
+```
+
+This adds a terminal overlay menu (Ctrl+P) and enables browser-to-terminal messaging.
+
+## Core Features
+
+### Browser Superpowers
+
+Start a proxy and your agent gains eyes into the browser:
+
+```
+proxy {action: "start", id: "app", target_url: "http://localhost:3000"}
+```
+
+Now your agent can:
+- Take screenshots
+- Inspect any element
+- Audit accessibility
+- See what you clicked
+
+### Floating Indicator
+
+Every proxied page gets a floating bug icon. Click to:
+- Send messages to your agent
+- Take area screenshots
+- Select elements to log
+- Open sketch mode
+
+### Sketch Mode
+
+Draw directly on your UI:
+- Shapes: rectangles, circles, arrows, freehand
+- Wireframes: buttons, inputs, sticky notes
+- Save and send to agent instantly
+
+### Real-Time Error Capture
+
+JavaScript errors automatically captured and available to your agent - no more forgetting to mention them.
+
+## MCP Tools
+
+| Tool | Description |
+|------|-------------|
+| `detect` | Auto-detect project type and scripts |
+| `run` | Run scripts or commands |
+| `proc` | Manage processes |
+| `proxy` | Reverse proxy with instrumentation |
+| `proxylog` | Query traffic logs |
+| `currentpage` | View page sessions |
+| `daemon` | Manage background service |
+
+## Browser API
+
+The proxy injects `window.__devtool` with 50+ diagnostic functions:
+
+```javascript
+__devtool.screenshot('name')              // Capture screenshot
+__devtool.inspect('#element')             // Full element analysis
+__devtool.auditAccessibility()            // A11y audit
+__devtool.sketch.open()                   // Enter sketch mode
+__devtool.interactions.getLastClick()     // Last click details
+```
+
+## Configuration
+
+Create `.agnt.kdl` in your project:
+
+```kdl
+scripts {
+    dev {
+        command "npm"
+        args "run" "dev"
+        autostart true
+    }
+}
+
+proxies {
+    frontend {
+        target "http://localhost:3000"
+        autostart true
+    }
+}
+```
+
+## Documentation
+
+- [GitHub](https://github.com/standardbeagle/agnt)
+- [Full Docs](https://standardbeagle.github.io/agnt/)
+
+## License
+
+MIT

agnt-0.7.6/README.md

@@ -0,0 +1,156 @@
+# agnt
+
+**Give your AI coding agent browser superpowers.**
+
+agnt is a new kind of tool designed for the age of AI-assisted development. It acts as a bridge between your AI coding agent and the browser, extending what's possible during vibe coding sessions.
+
+## What Does It Do?
+
+When you're in the flow with Claude Code, Cursor, or other AI tools, agnt lets your agent:
+
+- **See what you see** - Screenshots, DOM inspection, visual debugging
+- **Hear from you directly** - Send messages from browser to agent
+- **Sketch ideas together** - Draw wireframes directly on your UI
+- **Debug in real-time** - Capture errors, network traffic, performance
+- **Extend its thinking window** - Structured data uses fewer tokens than descriptions
+
+## Installation
+
+```bash
+pip install agnt
+# or
+uv pip install agnt
+# or
+pipx install agnt
+```
+
+## Quick Start
+
+### As MCP Server (Claude Code, Cursor, etc.)
+
+Add to your MCP configuration:
+
+```json
+{
+  "mcpServers": {
+    "agnt": {
+      "command": "agnt",
+      "args": ["serve"]
+    }
+  }
+}
+```
+
+Or with uvx:
+
+```json
+{
+  "mcpServers": {
+    "agnt": {
+      "command": "uvx",
+      "args": ["agnt", "serve"]
+    }
+  }
+}
+```
+
+### As PTY Wrapper
+
+Wrap your AI coding tool with overlay features:
+
+```bash
+agnt run claude --dangerously-skip-permissions
+agnt run cursor
+agnt run aider
+```
+
+This adds a terminal overlay menu (Ctrl+P) and enables browser-to-terminal messaging.
+
+## Core Features
+
+### Browser Superpowers
+
+Start a proxy and your agent gains eyes into the browser:
+
+```
+proxy {action: "start", id: "app", target_url: "http://localhost:3000"}
+```
+
+Now your agent can:
+- Take screenshots
+- Inspect any element
+- Audit accessibility
+- See what you clicked
+
+### Floating Indicator
+
+Every proxied page gets a floating bug icon. Click to:
+- Send messages to your agent
+- Take area screenshots
+- Select elements to log
+- Open sketch mode
+
+### Sketch Mode
+
+Draw directly on your UI:
+- Shapes: rectangles, circles, arrows, freehand
+- Wireframes: buttons, inputs, sticky notes
+- Save and send to agent instantly
+
+### Real-Time Error Capture
+
+JavaScript errors automatically captured and available to your agent - no more forgetting to mention them.
+
+## MCP Tools
+
+| Tool | Description |
+|------|-------------|
+| `detect` | Auto-detect project type and scripts |
+| `run` | Run scripts or commands |
+| `proc` | Manage processes |
+| `proxy` | Reverse proxy with instrumentation |
+| `proxylog` | Query traffic logs |
+| `currentpage` | View page sessions |
+| `daemon` | Manage background service |
+
+## Browser API
+
+The proxy injects `window.__devtool` with 50+ diagnostic functions:
+
+```javascript
+__devtool.screenshot('name')              // Capture screenshot
+__devtool.inspect('#element')             // Full element analysis
+__devtool.auditAccessibility()            // A11y audit
+__devtool.sketch.open()                   // Enter sketch mode
+__devtool.interactions.getLastClick()     // Last click details
+```
+
+## Configuration
+
+Create `.agnt.kdl` in your project:
+
+```kdl
+scripts {
+    dev {
+        command "npm"
+        args "run" "dev"
+        autostart true
+    }
+}
+
+proxies {
+    frontend {
+        target "http://localhost:3000"
+        autostart true
+    }
+}
+```
+
+## Documentation
+
+- [GitHub](https://github.com/standardbeagle/agnt)
+- [Full Docs](https://standardbeagle.github.io/agnt/)
+
+## License
+
+MIT

agnt-0.7.6/pyproject.toml

@@ -0,0 +1,69 @@
+[build-system]
+requires = ["hatchling"]
+build-backend = "hatchling.build"
+
+[project]
+name = "agnt"
+version = "0.7.6"
+description = "MCP server for AI coding agents - process management, reverse proxy with traffic logging, browser instrumentation, and sketch mode"
+readme = "README.md"
+license = "MIT"
+requires-python = ">=3.10"
+authors = [
+    { name = "Standard Beagle", email = "dev@standardbeagle.com" }
+]
+keywords = [
+    "mcp",
+    "model-context-protocol",
+    "ai",
+    "agent",
+    "development",
+    "devtools",
+    "proxy",
+    "debugging",
+    "frontend",
+    "diagnostics",
+    "sketch",
+    "wireframe",
+    "process-management",
+    "claude-code"
+]
+classifiers = [
+    "Development Status :: 4 - Beta",
+    "Environment :: Console",
+    "Intended Audience :: Developers",
+    "License :: OSI Approved :: MIT License",
+    "Operating System :: MacOS",
+    "Operating System :: Microsoft :: Windows",
+    "Operating System :: POSIX :: Linux",
+    "Programming Language :: Python :: 3",
+    "Programming Language :: Python :: 3.10",
+    "Programming Language :: Python :: 3.11",
+    "Programming Language :: Python :: 3.12",
+    "Topic :: Software Development :: Debuggers",
+    "Topic :: Software Development :: Testing",
+    "Topic :: Internet :: Proxy Servers",
+]
+dependencies = [
+    "httpx>=0.25.0",
+]
+
+[project.urls]
+Homepage = "https://standardbeagle.github.io/agnt/"
+Documentation = "https://standardbeagle.github.io/agnt/"
+Repository = "https://github.com/standardbeagle/agnt"
+Issues = "https://github.com/standardbeagle/agnt/issues"
+Changelog = "https://github.com/standardbeagle/agnt/blob/main/CHANGELOG.md"
+
+[project.scripts]
+agnt = "agnt:main"
+
+[tool.hatch.build.targets.wheel]
+packages = ["src/agnt"]
+
+[tool.hatch.build.targets.sdist]
+include = [
+    "/src",
+    "/README.md",
+    "/LICENSE",
+]

agnt-0.7.6/src/agnt/__init__.py

@@ -0,0 +1,174 @@
+"""
+agnt - MCP server for AI coding agents
+
+This package provides the agnt binary for use as an MCP server,
+offering process management, reverse proxy with traffic logging,
+browser instrumentation, and sketch mode.
+
+Usage:
+    Add to your MCP client configuration:
+    {
+        "mcpServers": {
+            "agnt": {
+                "command": "agnt",
+                "args": ["serve"]
+            }
+        }
+    }
+
+    Or run with PTY wrapper:
+        agnt run claude --dangerously-skip-permissions
+
+For more information, see:
+    https://standardbeagle.github.io/agnt/
+"""
+
+__version__ = "0.7.6"
+__all__ = ["main", "get_binary_path", "run"]
+
+import os
+import platform
+import stat
+import subprocess
+import sys
+import tempfile
+from pathlib import Path
+from typing import Optional
+
+import httpx
+
+REPO = "standardbeagle/agnt"
+VERSION = __version__
+BINARY_NAME = "agnt"
+
+
+def get_platform() -> str:
+    """Get the platform name for download."""
+    system = platform.system().lower()
+    if system == "darwin":
+        return "darwin"
+    elif system == "linux":
+        return "linux"
+    elif system == "windows":
+        return "windows"
+    else:
+        raise RuntimeError(f"Unsupported platform: {system}")
+
+
+def get_arch() -> str:
+    """Get the architecture name for download."""
+    machine = platform.machine().lower()
+    if machine in ("x86_64", "amd64"):
+        return "amd64"
+    elif machine in ("arm64", "aarch64"):
+        return "arm64"
+    else:
+        raise RuntimeError(f"Unsupported architecture: {machine}")
+
+
+def get_binary_name() -> str:
+    """Get the binary name for the current platform."""
+    if platform.system().lower() == "windows":
+        return f"{BINARY_NAME}.exe"
+    return BINARY_NAME
+
+
+def get_cache_dir() -> Path:
+    """Get the cache directory for storing the binary."""
+    if platform.system().lower() == "windows":
+        base = Path(os.environ.get("LOCALAPPDATA", Path.home() / "AppData" / "Local"))
+    elif platform.system().lower() == "darwin":
+        base = Path.home() / "Library" / "Caches"
+    else:
+        base = Path(os.environ.get("XDG_CACHE_HOME", Path.home() / ".cache"))
+
+    cache_dir = base / "agnt" / VERSION
+    cache_dir.mkdir(parents=True, exist_ok=True)
+    return cache_dir
+
+
+def get_download_url() -> str:
+    """Get the download URL for the binary."""
+    plat = get_platform()
+    arch = get_arch()
+    ext = ".exe" if plat == "windows" else ""
+    return f"https://github.com/{REPO}/releases/download/v{VERSION}/{BINARY_NAME}-{plat}-{arch}{ext}"
+
+
+def get_binary_path() -> Path:
+    """Get the path to the agnt binary, downloading if necessary."""
+    cache_dir = get_cache_dir()
+    binary_path = cache_dir / get_binary_name()
+
+    if binary_path.exists():
+        return binary_path
+
+    # Download the binary
+    url = get_download_url()
+    print(f"Downloading agnt v{VERSION}...", file=sys.stderr)
+    print(f"  Platform: {get_platform()}", file=sys.stderr)
+    print(f"  Architecture: {get_arch()}", file=sys.stderr)
+
+    try:
+        with httpx.Client(follow_redirects=True, timeout=60.0) as client:
+            response = client.get(url)
+            response.raise_for_status()
+
+            # Write to temp file first, then move
+            with tempfile.NamedTemporaryFile(
+                delete=False, dir=cache_dir, suffix=".tmp"
+            ) as tmp:
+                tmp.write(response.content)
+                tmp_path = Path(tmp.name)
+
+            # Move to final location
+            tmp_path.rename(binary_path)
+
+            # Make executable on Unix
+            if platform.system().lower() != "windows":
+                binary_path.chmod(binary_path.stat().st_mode | stat.S_IXUSR | stat.S_IXGRP | stat.S_IXOTH)
+
+            print(f"Successfully installed agnt to {binary_path}", file=sys.stderr)
+
+    except httpx.HTTPError as e:
+        print(f"Failed to download agnt: {e}", file=sys.stderr)
+        print("", file=sys.stderr)
+        print("You can manually download the binary from:", file=sys.stderr)
+        print(f"  https://github.com/{REPO}/releases/tag/v{VERSION}", file=sys.stderr)
+        print("", file=sys.stderr)
+        print("Or build from source:", file=sys.stderr)
+        print("  git clone https://github.com/standardbeagle/agnt.git", file=sys.stderr)
+        print("  cd agnt", file=sys.stderr)
+        print("  make build", file=sys.stderr)
+        sys.exit(1)
+
+    return binary_path
+
+
+def run(args: Optional[list[str]] = None) -> subprocess.CompletedProcess:
+    """Run agnt with the given arguments."""
+    binary_path = get_binary_path()
+    return subprocess.run([str(binary_path)] + (args or []))
+
+
+def main() -> None:
+    """Entry point for the agnt command."""
+    binary_path = get_binary_path()
+    try:
+        # Replace current process with the binary
+        if platform.system().lower() == "windows":
+            # Windows doesn't support exec, so use subprocess
+            result = subprocess.run([str(binary_path)] + sys.argv[1:])
+            sys.exit(result.returncode)
+        else:
+            os.execv(str(binary_path), [str(binary_path)] + sys.argv[1:])
+    except FileNotFoundError:
+        print(f"Error: Binary not found at {binary_path}", file=sys.stderr)
+        sys.exit(1)
+    except PermissionError:
+        print(f"Error: Binary at {binary_path} is not executable", file=sys.stderr)
+        sys.exit(1)
+
+
+if __name__ == "__main__":
+    main()