SWE-AF — AI Agent 工作流中文教程

1. Requirements (local)

• Python 3.12+
• AgentField control plane (af)
• AI provider API key (Anthropic, OpenRouter, OpenAI, or Google)

Autonomous Build Spotlight

Rust-based Python compiler benchmark (built autonomously):

<details> <summary>Measurement methodology</summary>

Throughput comparison measures different execution models: CPython subprocess spawn (~19ms per call → ~52 ops/s) vs RustPython pre-warmed interpreter pool (in-process). This is the real-world tradeoff the system was built to optimize — replacing repeated subprocess invocations with a persistent pool for short-snippet execution.

</details>

Artifact trail includes 175 tracked autonomous agents across planning, coding, review, merge, and verification.

Details: examples/llm-rust-python-compiler-sonnet/README.md

Deploy with Railway (fastest)

One click deploys SWE-AF + AgentField control plane + PostgreSQL. Set two environment variables in Railway:

• CLAUDE_CODE_OAUTH_TOKEN — run claude setup-token in Claude Code CLI (uses Pro/Max subscription credits)
• GH_TOKEN — GitHub personal access token with repo scope for PR creation

Once deployed, trigger a build:

curl -X POST https://<control-plane>.up.railway.app/api/v1/execute/async/swe-planner.build \
  -H "Content-Type: application/json" \
  -H "X-API-Key: this-is-a-secret" \
  -d '{"input": {"goal": "Add JWT auth", "repo_url": "https://github.com/user/my-repo"}}'

2. Install

python3.12 -m venv .venv
source .venv/bin/activate
python -m pip install --upgrade pip
python -m pip install -e ".[dev]"

4. Trigger a build

```bash

What Happens In One Build

• Architecture is generated and reviewed before coding starts
• Issues are dependency-sorted and run in parallel across isolated worktrees
• Each issue gets dedicated coder, tester, and reviewer passes
• Failed issues trigger advisor-driven adaptation (split, re-scope, or escalate)
• Escalations trigger replanning of the remaining DAG
• End result is merged, integration-tested, and verified against acceptance criteria

<p align="center"> <img src="assets/archi.png" alt="SWE-AF architecture" width="100%" /> </p>

Typical runs spin up 400-500+ agent instances across planning, execution, QA, and verification. For larger DAGs and repeated adaptation/replanning cycles, SWE-AF can scale into the high hundreds to thousands of agent invocations in a single build.

Docker

```bash cp .env.example .env

GitHub Repo Workflow (Clone -> Build -> PR)

Pass repo_url instead of repo_path to let SWE-AF clone and open a PR after execution.

curl -X POST http://localhost:8080/api/v1/execute/async/swe-planner.build \
  -H "Content-Type: application/json" \
  -d @- <<'JSON'
{
  "input": {
    "repo_url": "https://github.com/user/my-project",
    "goal": "Add comprehensive test coverage",
    "config": {
      "runtime": "claude_code",
      "models": {
        "default": "sonnet",
        "coder": "opus",
        "qa": "opus"
      }
    }
  }
}
JSON

Requirements:

• GH_TOKEN in .env with repo scope
• Repo access for that token

Full build: plan -> execute -> verify

POST /api/v1/execute/async/swe-planner.build

Quick Start

Optional: web search

Coding and review agents can look up external documentation, library APIs, error messages, and version/deprecation status during a build. This is opt-in via two env vars on the deployment:

OPENCODE_ENABLE_EXA=1
EXA_API_KEY=...

When set, opencode's built-in websearch and webfetch tools become available to every reasoner running through the open runtime — the model decides when to use them based on the task. Get a key at exa.ai.

The coder reasoner additionally gets a brief restraint guideline appended to its system prompt, so a long coding loop doesn't rabbit-hole on searches it could answer by reading the codebase. No setup required beyond the env vars; the wiring inherits parent env naturally through agentfield's CLI harness.

This works on the open runtime (opencode). The Claude runtime uses Anthropic's first-party WebSearch/WebFetch and is currently not wired here — file an issue if you want it.

Optionally add GH_TOKEN for PR workflow

docker compose up -d

Submit a build:

With open-source runtime (set OPENROUTER_API_KEY in .env)

curl -X POST http://localhost:8080/api/v1/execute/async/swe-planner.build \ -H "Content-Type: application/json" \ -d @- <<'JSON' { "input": { "goal": "Add JWT auth", "repo_url": "https://github.com/user/my-repo", "config": { "runtime": "open_code", "models": { "default": "openrouter/minimax/minimax-m2.5" } } } } JSON

With Codex CLI runtime

curl -X POST http://localhost:8080/api/v1/execute/async/swe-planner.build \ -H "Content-Type: application/json" \ -d @- <<'JSON' { "input": { "goal": "Add JWT auth", "repo_url": "https://github.com/user/my-project", "config": { "runtime": "codex", "models": { "default": "gpt-5.3-codex" } } } } JSON

Fast mode with Codex CLI runtime

curl -X POST http://localhost:8080/api/v1/execute/async/swe-fast.build \ -H "Content-Type: application/json" \ -d @- <<'JSON' { "input": { "goal": "Add a focused bug fix", "repo_url": "https://github.com/user/my-project", "config": { "runtime": "codex", "models": { "default": "gpt-5.3-codex" } } } } JSON

Add your API key: ANTHROPIC_API_KEY, OPENROUTER_API_KEY, OPENAI_API_KEY, or GOOGLE_API_KEY

API Reference

<details> <summary><strong>Agent endpoints</strong></summary>

Core async endpoints (returns an execution_id immediately):

```bash