Skip to content

README.md

Latest commit

 

History

History
78 lines (58 loc) · 1.76 KB

README.md

File metadata and controls

78 lines (58 loc) · 1.76 KB
Raw

Tests

This repo has two layers of verification:

  • Rust checks: fmt, clippy, unit tests
  • proxy integration tests: real Claude-shaped requests against a running proxy

Quick Start

Against a running proxy:

CHUTES_TEST_API_KEY=cpk_your_key ./test.sh --all

Defaults:

  • PROXY_URL=http://127.0.0.1:8080
  • MODEL=zai-org/GLM-4.5-Air

CI-Style Local Run

Build the release binary first:

cargo build --release

Start the mock backend:

python3 tests/mock_openai_backend.py --port 8000

Start the proxy in another shell:

BACKEND_URL=http://127.0.0.1:8000/v1/chat/completions \
HOST_PORT=8080 \
target/release/claude_openai_proxy

Run the main suite:

CHUTES_TEST_API_KEY=test \
PROXY_URL=http://127.0.0.1:8080 \
./test.sh --ci --all

Run the feature scripts:

CHUTES_TEST_API_KEY=test PROXY_URL=http://127.0.0.1:8080 ./tests/test_non_streaming.sh
MODEL=deepseek-r1 CHUTES_TEST_API_KEY=test PROXY_URL=http://127.0.0.1:8080 ./tests/test_thinking.sh
CHUTES_TEST_API_KEY=test PROXY_URL=http://127.0.0.1:8080 ./tests/validate_claude_api.sh

What Is Covered

  • basic requests, conversations, and parallel requests
  • Claude Code request shapes
  • streaming SSE responses
  • stream: false JSON responses
  • tool use and tool results
  • multimodal image inputs
  • token counting
  • model 404 handling and case correction
  • thinking output for reasoning-capable models

Important Files

  • test.sh unified entry point for the core suite
  • tests/mock_openai_backend.py deterministic OpenAI-compatible backend used in CI and local repros
  • tests/test_non_streaming.sh verifies non-streaming success and Anthropic-style JSON errors
  • tests/validate_claude_api.sh checks response shape for Claude-facing compatibility