coproxy

OpenAI-compatible API proxy backed by GitHub Copilot.

[!WARNING] This is a reverse-engineered proxy of GitHub Copilot API. It is not supported by GitHub, and may break unexpectedly. Use at your own risk.

[!WARNING] GitHub Security Notice:
Excessive automated or scripted use of Copilot (including rapid or bulk requests, such as via automated tools) may trigger GitHub's abuse-detection systems.
You may receive a warning from GitHub Security, and further anomalous activity could result in temporary suspension of your Copilot access.

GitHub prohibits use of their servers for excessive automated bulk activity or any activity that places undue burden on their infrastructure.

Please review:

• GitHub Acceptable Use Policies
• GitHub Copilot Terms

Use this proxy responsibly to avoid account restrictions.

Installation

From crates.io

cargo install coproxy

From source

git clone https://github.com/hedonhermdev/coproxy.git
cd coproxy
cargo install --path .

Prebuilt binaries

Download from GitHub Releases. Available for Linux (x86_64, aarch64) and macOS (x86_64, aarch64).

Docker

docker run -it -p 8080:8080 ghcr.io/hedonhermdev/coproxy serve --host 0.0.0.0

On first run, the container will print a GitHub device login URL and code to stdout. Complete the auth flow in your browser, and the server will start automatically.

To persist credentials across restarts, mount a volume:

docker run -it -p 8080:8080 -v coproxy-data:/data ghcr.io/hedonhermdev/coproxy serve --host 0.0.0.0 --state-dir /data

Quick start

# First-time login (opens GitHub device flow)
coproxy auth login

# Start the proxy
coproxy serve --port 8080 --api-surface all

# Verify it works
curl http://127.0.0.1:8080/v1/models -H "Authorization: Bearer any-key"

Usage with OpenAI Python client

from openai import OpenAI

client = OpenAI(
    base_url="http://127.0.0.1:8080/v1",
    api_key="any-key",  # or your --api-key value
)

# Chat completions
response = client.chat.completions.create(
    model="gpt-4o",
    messages=[{"role": "user", "content": "Hello!"}],
)
print(response.choices[0].message.content)

# Streaming
stream = client.chat.completions.create(
    model="gpt-4o",
    messages=[{"role": "user", "content": "Hello!"}],
    stream=True,
)
for chunk in stream:
    if chunk.choices[0].delta.content:
        print(chunk.choices[0].delta.content, end="")

# Responses API (model-dependent, e.g. gpt-5.4)
response = client.responses.create(
    model="gpt-5.4",
    input="Explain quantum computing in one sentence.",
)
print(response.output_text)

# List models
for model in client.models.list():
    print(model.id)

API surface

Configuration

CLI flags

Environment variables

Authentication

On first run, coproxy triggers GitHub device flow:

1. Prints a verification URL and user code.
2. Polls GitHub OAuth until you authorize.
3. Exchanges GitHub token for a GHCP API token.
4. Caches both tokens locally with restricted file permissions (0600).

coproxy auth login    # Interactive login
coproxy auth status   # Show cached token info
coproxy auth logout   # Remove cached credentials

Or skip device flow entirely by providing a GitHub token:

export GHCP_GITHUB_TOKEN=ghp_xxxxx
coproxy serve

Running as a service

systemd (Linux)

Copy contrib/coproxy.service to /etc/systemd/system/ and adjust as needed:

sudo cp contrib/coproxy.service /etc/systemd/system/
sudo systemctl daemon-reload
sudo systemctl enable --now coproxy

launchd (macOS)

Copy contrib/com.coproxy.plist to ~/Library/LaunchAgents/:

cp contrib/com.coproxy.plist ~/Library/LaunchAgents/
launchctl load ~/Library/LaunchAgents/com.coproxy.plist

Development

cargo fmt --all -- --check    # Format check
cargo clippy --all-targets    # Lint
cargo check                   # Type check
cargo test                    # Rust tests

# OpenAI compatibility tests (requires uv + GHCP auth)
scripts/run-openai-compat-tests.sh

License

MIT