detect_agent

This is a Python port of Vercel's detect-agent package.

A lightweight utility for detecting if code is being executed by an AI agent or automated development environment.

Installation

uv add detect_agent

Usage

from detect_agent import determine_agent

result = determine_agent()

if result["is_agent"]:
    agent = result["agent"]
    print(f"Running in {agent['name']} environment")
    # Adapt behavior for AI agent context

Agent Definitions

Detection rules are vendored from upstream agents.json (validated against agents.schema.json) and evaluated in array order; the first match wins.

Supported Agents

This package can detect the following AI agents and development environments:

• Cursor (Anysphere)
• Claude Code (Anthropic)
• Claude Cowork (Anthropic)
• Devin (Cognition Labs)
• Gemini CLI (Google)
• Codex (OpenAI)
• Antigravity (Google DeepMind)
• Augment
• Cline
• OpenCode
• OpenClaw
• Goose (Block)
• Junie (JetBrains)
• Kiro (AWS)
• Pi
• GitHub Copilot
• Replit
• Custom agents (via the AI_AGENT environment variable)

See agents.json for the exact environment variables and conditions used to detect each one.

The AI_AGENT Standard

We're promoting AI_AGENT as a universal environment variable standard for AI development tools. This allows any tool or library to easily detect when it's running in an AI-driven environment.

For AI Tool Developers

Set the AI_AGENT environment variable to identify your tool:

export AI_AGENT="your-tool-name"
# or
AI_AGENT="your-tool-name" your-command

Recommended Naming Convention

• Use lowercase with hyphens for multi-word names
• Include version information if needed, separated by an @ symbol
• Examples: claude-code, cursor-cli, devin@1, [email protected]

Development

uv sync --extra dev
uv run pytest
# Lint and format check (CI)
uv run ruff check . && uv run ruff format --check .
# Fix and format
uv run ruff check . --fix && uv run ruff format .

Use Cases

Adaptive Behavior

import os

from detect_agent import determine_agent


def setup_environment():
    result = determine_agent()

    if result["is_agent"]:
        agent = result["agent"]
        # Running in AI environment - adjust behavior
        os.environ["LOG_LEVEL"] = "verbose"
        print(f"Detected AI agent: {agent['name']}")

Telemetry and Analytics

import time

from detect_agent import determine_agent


def track_usage(event: str):
    result = determine_agent()

    analytics.track(
        event,
        {
            "agent": result["agent"]["name"] if result["is_agent"] else "human",
            "timestamp": time.time(),
        },
    )

Feature Toggles

from detect_agent import determine_agent


def should_enable_feature(feature: str) -> bool:
    result = determine_agent()

    # Enable experimental features for AI agents
    if result["is_agent"] and feature == "experimental-ai-mode":
        return True

    return False

Contributing

Adding New Agent Support

To add support for a new AI agent:

1. Sync detection rules from upstream agents.json into detect_agent/agents.json. Agents are evaluated in array order — the first match wins.
2. Sync or extend test cases in tests/testcases.json.
3. Update this README with the new agent information.

Links

• GitHub Repository
• Vercel upstream package