agentic-api

Stateful API logic for agentic applications using vLLM

A Rust-first project that is migrating agentic gateway functionality from Python into native Rust components. The previous Python gateway implementation has been removed. Design and migration decisions are tracked in the ADRs under docs/adr/.

Repository layout

• Rust source: src/
• Rust package manifest: Cargo.toml
• Documentation: docs/

Build

cargo build

Test

cargo test

Web search

The stateful /v1/responses executor supports OpenAI-compatible web_search tool declarations by normalizing them into a web_search function call for vLLM. Set YOU_API_KEY and YOU_API_BASE_URL to enable execution through You.com's Search API.

Using agentic-api with Codex

1. Start a vLLM server using any recipe from recipes.vllm.ai, for example:

   vllm serve Qwen/Qwen3-30B-A3B-FP8 --tool-call-parser qwen3_coder --enable-auto-tool-choice --reasoning-parser qwen3 --port 5050

2. Run agentic-api, pointing it at the vLLM server. Set YOU_API_KEY and YOU_API_BASE_URL to also enable web search:

   YOU_API_KEY=<your-you.com-api-key> YOU_API_BASE_URL=<you.com-api-base-url> \
     cargo run -p agentic-server -- --llm-api-base http://0.0.0.0:5050

3. Configure Codex to use agentic-api as a model provider. Create a config.toml in <codex-home-path>/.codex:

   [model_providers.agentic-api]
   name = "agentic-api"
   base_url = "http://localhost:9000/v1" # point to the agentic-api gateway url
   wire_api = "responses"
   requires_openai_auth = false
   supports_websockets = true

4. Run Codex against agentic-api:

   CODEX_HOME=/codex-home-path/.codex codex --disable image_generation -c model_provider=agentic-api -m Qwen/Qwen3-30B-A3B-FP8

Lint and format

cargo clippy --all-targets -- -D warnings
cargo fmt -- --check

Documentation

uv venv
uv pip install -r docs/requirements.txt
uv run mkdocs serve