Skip to content

Commit 984ae9f

Browse files
author
GitHub Expert
committed
fix: address review comments for DeepSeek configuration PR
- Move documentation to correct location: docs/providers/deepseek.md - Convert all documentation from Chinese to English - Fix model configuration: remove unsupported 'reasoning: true' for deepseek-chat - Update deprecated model IDs: deepseek-coder → deepseek-coder-v2 - Fix JSON format issues in documentation examples - Ensure all code comments are in English - Merge security fixes from security-fix-20260304 branch
2 parents e286f45 + 3322828 commit 984ae9f

20 files changed

+2980
-6
lines changed

‎docs/deepseek-config-example.md.backup‎

Lines changed: 157 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,157 @@
1+
# DeepSeek 大模型配置指南
2+
3+
## 概述
4+
5+
本文档介绍如何在 OpenClaw 中配置 DeepSeek 大模型。DeepSeek 是一个强大的开源大语言模型，支持多种任务和功能。
6+
7+
## 环境变量配置
8+
9+
在 `.env` 文件中添加以下配置：
10+
11+
```bash
12+
# DeepSeek API 密钥
13+
DEEPSEEK_API_KEY=your-deepseek-api-key-here
14+
```
15+
16+
## OpenClaw 配置文件示例
17+
18+
在 `openclaw.json` 配置文件中添加 DeepSeek 模型提供者：
19+
20+
```json
21+
{
22+
"models": {
23+
"mode": "merge",
24+
"providers": {
25+
"deepseek": {
26+
"baseUrl": "https://api.deepseek.com",
27+
"apiKey": {
28+
"env": "DEEPSEEK_API_KEY"
29+
},
30+
"api": "openai-completions",
31+
"models": [
32+
{
33+
"id": "deepseek-chat",
34+
"name": "DeepSeek Chat",
35+
"reasoning": true,
36+
"input": ["text"],
37+
"cost": {
38+
"input": 0.14,
39+
"output": 0.28,
40+
"cacheRead": 0.0,
41+
"cacheWrite": 0.0
42+
},
43+
"contextWindow": 128000,
44+
"maxTokens": 4096
45+
},
46+
{
47+
"id": "deepseek-coder",
48+
"name": "DeepSeek Coder",
49+
"reasoning": true,
50+
"input": ["text"],
51+
"cost": {
52+
"input": 0.14,
53+
"output": 0.28,
54+
"cacheRead": 0.0,
55+
"cacheWrite": 0.0
56+
},
57+
"contextWindow": 128000,
58+
"maxTokens": 4096
59+
}
60+
]
61+
}
62+
}
63+
}
64+
}
65+
```
66+
67+
## 配置说明
68+
69+
### 1. 基础配置
70+
- **baseUrl**: DeepSeek API 端点
71+
- **apiKey**: 从环境变量 `DEEPSEEK_API_KEY` 读取 API 密钥
72+
- **api**: 使用 OpenAI 兼容的 API 格式
73+
74+
### 2. 模型列表
75+
目前支持两个主要模型：
76+
77+
#### DeepSeek Chat
78+
- **ID**: `deepseek-chat`
79+
- **名称**: DeepSeek Chat
80+
- **推理能力**: 支持
81+
- **输入类型**: 文本
82+
- **上下文窗口**: 128K tokens
83+
- **最大输出**: 4096 tokens
84+
85+
#### DeepSeek Coder
86+
- **ID**: `deepseek-coder`
87+
- **名称**: DeepSeek Coder
88+
- **推理能力**: 支持
89+
- **输入类型**: 文本
90+
- **上下文窗口**: 128K tokens
91+
- **最大输出**: 4096 tokens
92+
93+
### 3. 成本配置
94+
- **输入**: $0.14 每百万 tokens
95+
- **输出**: $0.28 每百万 tokens
96+
- **缓存**: 免费
97+
98+
## 使用示例
99+
100+
### 在会话中指定模型
101+
```bash
102+
# 使用 DeepSeek Chat 模型
103+
openclaw --model deepseek/deepseek-chat
104+
105+
# 使用 DeepSeek Coder 模型
106+
openclaw --model deepseek/deepseek-coder
107+
```
108+
109+
### 在配置中设置默认模型
110+
```json
111+
{
112+
"defaultModel": "deepseek/deepseek-chat",
113+
"models": {
114+
// ... 模型配置
115+
}
116+
}
117+
```
118+
119+
## 注意事项
120+
121+
1. **API 密钥**: 需要从 DeepSeek 官网获取 API 密钥
122+
2. **速率限制**: 注意 API 调用速率限制
123+
3. **模型可用性**: 确保选择的模型在 API 中可用
124+
4. **成本控制**: 监控 API 使用成本
125+
126+
## 故障排除
127+
128+
### 常见问题
129+
130+
1. **认证失败**
131+
- 检查 `DEEPSEEK_API_KEY` 环境变量是否正确设置
132+
- 验证 API 密钥是否有效
133+
134+
2. **模型不可用**
135+
- 检查模型 ID 是否正确
136+
- 确认 API 端点是否支持该模型
137+
138+
3. **网络问题**
139+
- 检查网络连接
140+
- 验证 API 端点可访问性
141+
142+
### 调试命令
143+
```bash
144+
# 检查环境变量
145+
echo $DEEPSEEK_API_KEY
146+
147+
# 测试 API 连接
148+
curl -H "Authorization: Bearer $DEEPSEEK_API_KEY" \
149+
https://api.deepseek.com/v1/models
150+
```
151+
152+
## 更新日志
153+
154+
### 2026-03-04
155+
- 初始版本：添加 DeepSeek 配置支持
156+
- 支持 DeepSeek Chat 和 DeepSeek Coder 模型
157+
- 添加完整的配置示例和文档

‎docs/providers/deepseek.md‎

Lines changed: 162 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,162 @@
1+
# DeepSeek Model Configuration Guide
2+
3+
## Overview
4+
5+
This document explains how to configure DeepSeek large language models in OpenClaw. DeepSeek is a powerful open-source large language model that supports various tasks and functionalities.
6+
7+
## Environment Variables Configuration
8+
9+
Add the following configuration to your `.env` file:
10+
11+
```bash
12+
# DeepSeek API key
13+
DEEPSEEK_API_KEY=your-deepseek-api-key-here
14+
```
15+
16+
## OpenClaw Configuration Example
17+
18+
Add DeepSeek model provider to your `openclaw.json` configuration file:
19+
20+
```json
21+
{
22+
"modelProviders": {
23+
"deepseek": {
24+
"kind": "openai",
25+
"baseURL": "https://api.deepseek.com",
26+
"apiKey": "${DEEPSEEK_API_KEY}",
27+
"defaultModel": "deepseek-chat"
28+
}
29+
}
30+
}
31+
```
32+
33+
## Available DeepSeek Models
34+
35+
### 1. DeepSeek Chat Models
36+
- `deepseek-chat`: General purpose chat model (maps to DeepSeek-V3)
37+
- `deepseek-chat-v3`: Latest version of DeepSeek chat model
38+
39+
**Note**: The `deepseek-chat` model does not support the `reasoning: true` parameter. For reasoning capabilities, use the specific reasoning models mentioned below.
40+
41+
### 2. DeepSeek Coder Models
42+
- `deepseek-coder-v2`: Latest code generation model (recommended)
43+
- `deepseek-coder`: Legacy model (deprecated, use `deepseek-coder-v2` instead)
44+
45+
### 3. DeepSeek Reasoning Models
46+
- `deepseek-reasoner`: Specialized model for complex reasoning tasks
47+
- `deepseek-math`: Optimized for mathematical reasoning
48+
49+
## Model Configuration Examples
50+
51+
### Basic Chat Configuration
52+
```json
53+
{
54+
"models": {
55+
"deepseek-default": {
56+
"provider": "deepseek",
57+
"model": "deepseek-chat",
58+
"temperature": 0.7,
59+
"maxTokens": 2048
60+
}
61+
}
62+
}
63+
```
64+
65+
### Code Generation Configuration
66+
```json
67+
{
68+
"models": {
69+
"deepseek-coder": {
70+
"provider": "deepseek",
71+
"model": "deepseek-coder-v2",
72+
"temperature": 0.2,
73+
"maxTokens": 4096
74+
}
75+
}
76+
}
77+
```
78+
79+
### Reasoning Task Configuration
80+
```json
81+
{
82+
"models": {
83+
"deepseek-reasoning": {
84+
"provider": "deepseek",
85+
"model": "deepseek-reasoner",
86+
"temperature": 0.3,
87+
"maxTokens": 4096
88+
}
89+
}
90+
}
91+
```
92+
93+
## Usage Examples
94+
95+
### 1. Agent Configuration
96+
```json
97+
{
98+
"agents": {
99+
"coding-assistant": {
100+
"model": "deepseek-coder",
101+
"systemPrompt": "You are an expert coding assistant..."
102+
},
103+
"general-assistant": {
104+
"model": "deepseek-default",
105+
"systemPrompt": "You are a helpful AI assistant..."
106+
}
107+
}
108+
}
109+
```
110+
111+
### 2. Skill Configuration
112+
```json
113+
{
114+
"skills": {
115+
"code-review": {
116+
"model": "deepseek-coder",
117+
"parameters": {
118+
"temperature": 0.1,
119+
"maxTokens": 1024
120+
}
121+
}
122+
}
123+
}
124+
```
125+
126+
## Best Practices
127+
128+
1. **API Key Security**: Always use environment variables for API keys
129+
2. **Model Selection**: Choose the appropriate model for your specific task
130+
3. **Error Handling**: Implement proper error handling for API failures
131+
4. **Rate Limiting**: Respect DeepSeek API rate limits
132+
5. **Cost Management**: Monitor token usage for cost control
133+
134+
## Troubleshooting
135+
136+
### Common Issues
137+
138+
1. **Authentication Errors**: Verify your `DEEPSEEK_API_KEY` is correctly set
139+
2. **Model Not Found**: Ensure you're using the correct model identifier
140+
3. **Rate Limit Exceeded**: Implement exponential backoff for retries
141+
4. **Network Issues**: Check your internet connection and firewall settings
142+
143+
### Debugging Tips
144+
145+
- Enable debug logging in OpenClaw configuration
146+
- Check the OpenClaw logs for detailed error messages
147+
- Verify network connectivity to `api.deepseek.com`
148+
- Test with simple prompts to isolate issues
149+
150+
## Additional Resources
151+
152+
- [DeepSeek Official Documentation](https://platform.deepseek.com/api-docs/)
153+
- [OpenClaw Model Providers Documentation](https://docs.openclaw.ai/guides/model-providers)
154+
- [API Rate Limits and Pricing](https://platform.deepseek.com/pricing)
155+
156+
## Support
157+
158+
For issues with DeepSeek integration:
159+
1. Check the OpenClaw documentation
160+
2. Review the DeepSeek API documentation
161+
3. Open an issue on the OpenClaw GitHub repository
162+
4. Contact DeepSeek support for API-specific issues

‎extensions/diffs/assets/viewer-runtime.js‎

Lines changed: 1 addition & 1 deletion
Some generated files are not rendered by default. Learn more about customizing how changed files appear on GitHub.

‎extensions/feishu/src/send.ts‎

Lines changed: 4 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -8,6 +8,8 @@ import { getFeishuRuntime } from "./runtime.js";
88
import { assertFeishuMessageApiSuccess, toFeishuSendResult } from "./send-result.js";
99
import { resolveFeishuSendTarget } from "./send-target.js";
1010
import type { FeishuSendResult } from "./types.js";
11+
// Security fix: Import safe JSON parsing function
12+
import { validateSafeJson } from "../../../src/security/input-validation.js";
1113

1214
const WITHDRAWN_REPLY_ERROR_CODES = new Set([230011, 231003]);
1315

@@ -123,7 +125,8 @@ function parseQuotedMessageContent(rawContent: string, msgType: string): string
123125

124126
let parsed: unknown;
125127
try {
126-
parsed = JSON.parse(rawContent);
128+
// Security fix: Use safe JSON parsing function
129+
parsed = validateSafeJson(rawContent);
127130
} catch {
128131
return rawContent;
129132
}

0 commit comments

Comments
 (0)