CodingAgent-Reproduce/config.example.yaml at main · RavenCaffeine/CodingAgent-Reproduce · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
# ============================================================================
# MewCode config / 配置文件 —— copy to config.yaml / 复制为 config.yaml
#
# Keep exactly ONE provider block active (uncommented); comment out the rest.
# 只保留一组 provider 启用（顶格、无 #），其余整段注释掉。
# Switch provider = uncomment the one you want, comment the others.
# 切换 provider = 取消想用那组的注释，注释掉其他组。
#
# Four core fields / 四个核心字段:
#   protocol  protocol to use / 走哪家协议   anthropic | openai | deepseek
#   model     model id / 用哪个模型
#   base_url  endpoint (optional) / 请求地址（可选）
#   api_key   auth / 认证                    see "API key" note below / 见下方说明
# Optional / 可选:
#   thinking        extended thinking / 扩展思考  Anthropic & DeepSeek; OpenAI ignores
#   name            display label / 显示名
#   context_window  auto-compaction window / 自动压缩窗口   default 200000 / 默认 200000
#
# ── API key: 3 ways / 三种写法 (pick one / 任选其一) ─────────────────────────
#   1) manual literal / 手动填入:  api_key: sk-xxxxx
#   2) env import / 环境导入:      api_key: ${OPENAI_API_KEY}
#   3) explicit env / 显式声明:    api_key_env: OPENAI_API_KEY
#   If omitted / 都不写: falls back to the provider's default env var
#     (ANTHROPIC_API_KEY / OPENAI_API_KEY / DEEPSEEK_API_KEY).
#   Priority / 优先级: manual literal > env var / 手动填入 > 环境变量
# ============================================================================


# ---------------------------------------------------------------------------
# Option A / 方案 A: Anthropic Claude
#   uncomment this block & comment out the others / 取消注释并注释掉其他组
# ---------------------------------------------------------------------------
# protocol: anthropic
# model: claude-sonnet-4-6             # e.g. claude-opus-4-6 / claude-sonnet-4-6
# base_url: https://api.anthropic.com  # optional / 可选
# api_key: ${ANTHROPIC_API_KEY}        # or literal / 或明文: sk-ant-xxxxx
# thinking: true                       # adaptive on opus/sonnet-4-6+ / 否则固定 budget
# context_window: 200000               # ch08: auto-compaction threshold base / 自动压缩阈值的窗口


# ---------------------------------------------------------------------------
# Option B / 方案 B: OpenAI (Responses API)  (active / 当前启用)
# ---------------------------------------------------------------------------
protocol: openai
model: gpt-4.1                       # e.g. gpt-4.1 / gpt-4o / o4-mini
base_url: https://api.openai.com/v1  # optional / 可选
api_key: ${OPENAI_API_KEY}           # or literal / 或明文: sk-xxxxx
thinking: false                      # OpenAI ignores / OpenAI 忽略此项
context_window: 200000               # ch08: auto-compaction threshold base /
                                     # 自动压缩阈值的窗口；预算小（如紧的 TPM 限制）就调小


# ---------------------------------------------------------------------------
# Option C / 方案 C: DeepSeek V4 (Chat Completions API)
#   uncomment this block & comment out the others / 取消注释并注释掉其他组
# ---------------------------------------------------------------------------
# protocol: deepseek
# model: deepseek-v4-pro              # deepseek-v4-pro / deepseek-v4-flash
# base_url: https://api.deepseek.com  # optional, this is the default / 可选，默认值
# api_key: ${DEEPSEEK_API_KEY}        # or literal / 或明文: sk-xxxxx
# thinking: false                     # reasoning streams auto / reasoning 自动流式
# context_window: 128000              # ch08: auto-compaction threshold base / 自动压缩阈值的窗口


# ---------------------------------------------------------------------------
# MCP servers / 外部 MCP 服务器 (optional / 可选, ch07)
#   Connected at startup; each server's tools appear as mcp_<server>_<tool>.
#   启动时连接；每个 server 的工具以 mcp_<server>_<tool> 形式注册。
#   Declare exactly ONE transport per server / 每个 server 只能选一种传输:
#     stdio:  command + args (+ env)   |   http: url (+ headers)
#   headers/env values support ${VAR} expansion / 支持 ${VAR} 环境变量展开
# ---------------------------------------------------------------------------
# mcp_servers:
#   context7:                          # stdio: local subprocess / 本地子进程
#     command: npx
#     args: ["-y", "@upstash/context7-mcp"]
#   remote:                            # http: Streamable HTTP / 远程
#     url: https://example.com/mcp
#     headers:
#       Authorization: "Bearer ${MCP_TOKEN}"