Repository navigation
Expand file tree
/
Copy pathsettings.example.json
More file actions
67 lines (67 loc) · 2.15 KB
/
Copy pathsettings.example.json
File metadata and controls
67 lines (67 loc) · 2.15 KB
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
{
"base_url": "https://api.openai.com/v1",
"api_key": "<你的密钥>",
"model": "gpt-4o-mini",
"providers": [
{
"name": "deepseek",
"base_url": "https://api.deepseek.com/v1",
"api_key_env": "DEEPSEEK_API_KEY"
},
{
"name": "ollama",
"base_url": "http://localhost:11434/v1"
},
{
"name": "glm",
"base_url": "https://open.bigmodel.cn/api/paas/v4",
"api_key": "<你的密钥>",
"thinking_style": "thinking"
}
],
"groups": [
{
"name": "gpt",
"provider": "default",
"models": [
"gpt-4o",
"gpt-4o-mini"
]
},
{
"name": "ds",
"provider": "deepseek",
"models": [
"deepseek-v4-pro",
"deepseek-v4-flash"
]
},
{
"name": "local",
"provider": "ollama",
"models": [
"qwen2.5"
]
},
{
"name": "glm",
"provider": "glm",
"models": [
"glm-4.5"
]
}
],
"system_prompt": "You are Ember, a software-engineering agent inside a terminal. You help the user build and fix software by inspecting the codebase and running real tools.\nWork habits:\n- Be small-step: inspect before you act, act before you claim.\n- After changing code, run the project's build and tests to verify; if they fail, read the failure, fix, and re-verify. Never repeat the same wrong move twice.\n- Prefer dedicated tools (file_read/file_write/edit/patch/grep/glob) over shell for file work; use shell_exec for builds, tests, git and anything that runs a program.\n- Use background_start for long-running work and check background_status when asked, you will also be notified when it finishes.\n- Read files before editing them so your edits match exact text. State what you changed and one line of evidence (e.g. a test that now passes).\n- When stuck, ask the user (ask_user) instead of guessing.",
"theme": "dark",
"mode": "standard",
"thinking": "medium",
"budget": 6000,
"tool_result_cap": 4096,
"prune_before_send": true,
"json_mode": false,
"shell_enabled": true,
"cost_in_per_m": 0,
"cost_out_per_m": 0,
"cost_budget_usd": 10,
"workspace_guard": true
}