Hivulse蜂巢AI-Gen-Tech Docs 自动生成技术文档
v0.1.4hivulse蜂巢 AI 是一款面向软件开发的自动化技术文档生成工具,通过指定目录代码一键生成多种规范化技术文档。目前已支持的文档类型包括:用户需求说明书、需求规格说明书、系统概要设计说明、系统详细设计说明等10几种报告。申请API Key请访问 www.hivulse.com
Security Scan
OpenClaw
Benign
medium confidencePurpose & Capability
The skill name/description (automated technical-document generation from a project directory) matches what the code does: it collects files, filters common directories, uploads files to a hivulse API, checks status, and requests document generation. Requiring an HIVULSE_API_KEY is appropriate for this remote API integration.
Instruction Scope
Runtime instructions and code focus on the documented workflow (gather files, upload, request generation). However the tool will upload nearly all files in the specified directory (only node_modules, venv, .git, __pycache__, .idea, .vscode and *.pyc/*.log are explicitly excluded). It does not filter .env, *.pem, keys, or other secret files — so sensitive files in the project directory would be uploaded to the remote service. The code also reads OpenClaw configuration to find an API key (purpose-aligned), but one implementation path contains a hard-coded developer path (/Users/superlk/.openclaw/openclaw.json), which looks like a leftover and is inconsistent with other code that uses Path.home().
Install Mechanism
There is no remote installer or downloads; this is an instruction + Python script package with a small requirements.txt (requests, pathlib2). No external arbitrary download URLs or archive extraction are used in the provided files.
Credentials
The skill only requires a single credential (HIVULSE_API_KEY), which is proportional to contacting a remote API. The code attempts to source that key from several places (OpenClaw config, environment, local ~/.hivulseai/config.json). Reading the OpenClaw config is reasonable for convenience, but the code will read a file in the user's home directory (potentially exposing other configuration data during debugging/logging). The number of env vars requested is minimal and appropriate.
Persistence & Privilege
The skill does not request always:true and does not attempt to modify other skills or system-wide settings. It creates/uses a local config directory (~/.hivulseai) for storing its API key and last-used directory — this is normal for a CLI tool.
Assessment
This skill appears to do what it claims (upload your project and call hivulse cloud APIs) and only needs an HIVULSE_API_KEY. Before installing or using it: 1) Audit the code or run in a sandbox: the tool will upload nearly all files under the directory you give it — ensure you do not upload secrets (environment files, private keys, credentials, .env, .pem, etc.), since those are not excluded by default. 2) Confirm the API endpoint and privacy policy for the hivulse service (the code uses 'https://cloud.hivulse.com' but other messages reference localhost — clarify which endpoint will be used). 3) Note minor sloppy items (a hard-coded path /Users/superlk in one code path and inconsistent endpoint mentions) — these look like developer leftovers rather than malicious behavior but justify checking the files before use. 4) If you must use on sensitive code, strip sensitive files from the project or run a filtered copy. 5) If you have low trust in the remote service, do not provide the API key or upload private repositories.Like a lobster shell, security has layers — review code before you run it.
Runtime requirements
📄 Clawdis
EnvHIVULSE_API_KEY
Primary envHIVULSE_API_KEY
latest
hivulseAI - 自动化技术文档生成工具
通过API接口将指定目录的文件上传并生成各种技术文档的完整自动化工具。
🚀 快速开始
环境变量设置
export HIVULSE_API_KEY="your-api-key-here"
📋 支持的文档类型
| 编号 | 文档类型 | 描述 |
|---|---|---|
| 19 | 用户需求说明书 | 用户需求分析文档 |
| 2 | 需求规格说明书 | 详细需求规格 |
| 4 | 系统概要设计说明 | 系统架构设计 |
| 5 | 系统详细设计说明 | 详细设计文档 |
| 8 | 软件单元测试计划 | 单元测试计划 |
| 9 | 软件单元测试用例 | 单元测试用例 |
| 10 | 软件单元测试报告 | 单元测试报告 |
| 1 | 系统测试计划 | 系统测试计划 |
| 12 | 系统测试用例 | 系统测试用例 |
| 15 | 系统测试报告 | 系统测试报告 |
| 13 | 网络安全漏洞自评报告 | 安全评估报告 |
| 20 | 软件用户测试用例 | 用户测试用例 |
| 21 | 软件用户测试报告 | 用户测试报告 |
🎯 触发关键词
在OpenClaw会话中,当用户提到以下关键词时会自动触发此技能:
- "生成技术文档"
- "创建文档"
- "文档生成"
- "hivulseAI"
- "上传代码生成文档"
- "需求规格说明书"
- "系统设计文档"
🔧 OpenClaw技能调用方式
方式1:直接触发
在对话中直接使用触发关键词,例如:
- "帮我生成技术文档"
- "使用hivulseAI生成需求规格说明书"
- "为这个项目创建系统设计文档"
📋 参数说明
| 参数 | 说明 | 示例 |
|---|---|---|
| 项目路径 | 要生成文档的代码项目目录 | /Users/project/myapp |
| 文档类型 | 文档类型编号(2-21) | 2(需求规格说明书) |
| 任务名称 | 可选,文档任务名称 | "项目需求分析" |
🌐 API配置
- API密钥:通过配置文件设置,无需环境变量
配置文件位置:
~/.hivulseai/config.json
手动编辑配置:
{
"api_key": "your-api-key-here",
"last_used_directory": ""
}
📁 文件过滤规则
系统会自动过滤以下目录和文件:
node_modules- Node.js依赖目录venv- Python虚拟环境.git- Git版本控制__pycache__- Python缓存.idea- IDE配置.vscode- VS Code配置*.pyc- Python编译文件*.log- 日志文件
🔄 API调用流程
1. 文件上传阶段
POST /api/v1/repo/upload/file/
- 第一个文件:不带branch_id
- 后续文件:带branch_id(从第一个文件响应中获取)
2. 状态检查阶段
POST /api/v1/repo/upload/status/
- 参数:uuid(使用default_branch_id)
3. 文档生成阶段
POST /api/v1/template_wiki/
- 参数:task_name, template_base_id, branch_id, repo_id, is_advanced
⚠️ 注意事项
- API密钥安全:确保
HIVULSE_API_KEY环境变量正确设置 - 网络连接:确保可以访问指定的API地址
- 文件权限:确保对目标目录有读取权限
- 文件大小:注意单个文件大小限制
- 备份重要文件:建议先备份重要代码
🛠️ 故障排除
常见错误
- "API密钥未设置" → 检查
HIVULSE_API_KEY环境变量 - "目录不存在" → 检查路径是否正确
- "上传失败" → 检查网络连接和API服务状态
- "文档类型不支持" → 检查文档类型编号
📞 支持
如有问题,请检查:
- API服务是否正常运行
- 环境变量是否正确设置
- 网络连接是否正常
- 文件路径和权限是否正确
Comments
Loading comments...
