← 返回 Skills 市场
137
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install ai-debug-assistant
功能描述
AI-powered debugging assistant. Analyze error logs, explain error messages, parse stack traces, and get fix suggestions with cheatsheets for 8 languages. Pow...
使用说明 (SKILL.md)
Debug Assistant
AI-powered debugging from your terminal. Analyze error logs, explain error messages, parse stack traces, get fix suggestions, and access debugging cheatsheets for 8 languages.
Powered by Evolink.ai
When to Use
- User shares an error log and asks "what went wrong?"
- User pastes an error message and wants an explanation
- User has a stack trace and needs help understanding it
- User has buggy code and a known error, wants fix suggestions
- User needs quick debugging commands for a specific language
Quick Start
1. Set your EvoLink API key
export EVOLINK_API_KEY="your-key-here"
Get a free key: evolink.ai/signup
2. Analyze an error
bash scripts/debug.sh analyze error.log
bash scripts/debug.sh explain "Cannot read property 'map' of undefined"
3. Get a cheatsheet
bash scripts/debug.sh cheatsheet python
Capabilities
AI Commands (require EVOLINK_API_KEY)
| Command | Description |
|---|---|
analyze \x3Cfile> |
Analyze error log — summary, root cause, error chain, fix steps |
explain \x3Cerror_message> |
Explain error — meaning, common causes, quick fix, proper fix |
trace \x3Cfile> |
Parse stack trace — origin, call chain, root cause, fix |
suggest \x3Cfile> --error \x3Cmsg> |
Fix suggestions — bug location, why it fails, corrected code |
Info Commands (no API key needed)
| Command | Description |
|---|---|
languages |
List all supported languages |
cheatsheet [language] |
Debugging commands and common errors |
Supported Languages
| Language | Debugger | Key Tools |
|---|---|---|
javascript |
Node --inspect, Chrome DevTools | console.trace, memory profiling |
python |
pdb, breakpoint() | tracemalloc, cProfile |
go |
Delve | -race flag, pprof |
rust |
rust-gdb, rust-lldb | RUST_BACKTRACE, env_logger |
java |
Remote JDWP | jmap, jstack, GC logging |
network |
curl -v, dig | lsof, netstat, ss |
css |
Outline trick | Grid/Flex inspectors |
git |
git bisect | Automated bisect with test scripts |
Examples
Analyze an error log
bash scripts/debug.sh analyze server-crash.log
Output:
**Error Summary:** Node.js process crashed due to unhandled promise rejection
in the database connection pool.
**Root Cause:** The PostgreSQL connection string contains an expired SSL
certificate path, causing all new connections to fail silently...
**Fix Steps:**
1. Update the SSL certificate at /etc/ssl/certs/db.pem
2. ...
Explain an error message
bash scripts/debug.sh explain "ECONNREFUSED 127.0.0.1:5432"
Parse a stack trace
bash scripts/debug.sh trace crash-dump.txt
Suggest fixes for buggy code
bash scripts/debug.sh suggest api/handler.py --error "KeyError: 'user_id'"
Configuration
| Variable | Default | Required | Description |
|---|---|---|---|
EVOLINK_API_KEY |
— | Yes | Your EvoLink API key. Get one free |
EVOLINK_MODEL |
claude-opus-4-6 |
No | Model for AI analysis |
Required binaries: python3, curl
Security
Data Transmission
AI commands send user-provided content (error logs, messages, code files) to api.evolink.ai for analysis by Claude. By setting EVOLINK_API_KEY and using these commands, you consent to this transmission. Data is not stored after the response is returned. The languages and cheatsheet commands run entirely locally and never transmit data.
Network Access
api.evolink.ai— AI error analysis (AI commands only)
Persistence & Privilege
This skill creates temporary files for API payload construction which are cleaned up automatically. No credentials or persistent data are stored. The skill only reads files explicitly passed as arguments — it never scans or reads files on its own.
Links
安全使用建议
This skill appears to do what it says: AI commands send any logs, error messages, or code you provide to api.evolink.ai using the EVOLINK_API_KEY. Before installing or using it: 1) Do not send secrets, credentials, or sensitive PII in logs or files you submit to the AI endpoint. 2) Review the bundled scripts (scripts/debug.sh) yourself — installing via npx executes installer code locally. 3) Verify the privacy/security policy of evolink.ai if you need guarantees about retention or handling of submitted data. 4) If you need offline analysis, stick to the local info commands (languages, cheatsheet) which do not make network calls. 5) If you want stronger isolation, create a dedicated API key with limited scope (if supported) and rotate it if it is ever exposed.
功能分析
Type: OpenClaw Skill
Name: ai-debug-assistant
Version: 1.0.0
The skill is a legitimate AI-powered debugging assistant designed to analyze error logs and code snippets via the Evolink.ai API. While it transmits local file content to an external endpoint (api.evolink.ai), this behavior is explicitly disclosed in the SKILL.md and _meta.json documentation. The implementation in scripts/debug.sh uses standard practices for handling JSON payloads and temporary files, and there is no evidence of malicious intent, obfuscation, or unauthorized data exfiltration.
能力评估
Purpose & Capability
The skill is described as an AI debugging assistant and the scripts implement that: they read files passed as arguments, build a prompt, and post to api.evolink.ai. Required binaries (python3, curl) are actually used. The single required env var (EVOLINK_API_KEY) is the API credential needed to call the external service.
Instruction Scope
The SKILL.md and scripts explicitly send user-provided content (error logs, messages, code) to api.evolink.ai for AI analysis. The README/SKILL.md declare this. The scripts do not attempt to read arbitrary files or other environment variables beyond EVOLINK_API_KEY and optional EVOLINK_MODEL, and they use temporary files for payload construction (with a trap to remove them). This is scope-appropriate but means you should avoid sending secrets or PII in logs.
Install Mechanism
The package includes an npm installer (bin/install.js) that copies bundled skill-files into a skills/ directory and updates a .clawhub lock. There are no remote downloads or obscure URLs in the install step. Installing via npx will download the npm package from the registry — standard but worth noting because that runs installer code locally.
Credentials
Only EVOLINK_API_KEY is required (EVOLINK_MODEL optional). That single credential directly supports the stated purpose (calling the Evolink API). No unrelated credentials, system tokens, or config paths are requested.
Persistence & Privilege
The skill does not request always:true, does not modify other skills' configs, and does not persist credentials. Installer writes files into a project-level skills directory and updates a .clawhub lockfile — expected behavior for a skill installer. Temporary payload files are created and removed with a trap on exit.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install ai-debug-assistant - 安装完成后，直接呼叫该 Skill 的名称或使用
/ai-debug-assistant触发 - 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v1.0.0
Initial release: AI-powered debugging for 8 languages with error analysis, stack trace parsing, and fix suggestions
元数据
常见问题
Debug Assistant 是什么？
AI-powered debugging assistant. Analyze error logs, explain error messages, parse stack traces, and get fix suggestions with cheatsheets for 8 languages. Pow... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 137 次。
如何安装 Debug Assistant？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install ai-debug-assistant」即可一键安装，无需额外配置。
Debug Assistant 是免费的吗？
是的，Debug Assistant 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Debug Assistant 支持哪些平台？
Debug Assistant 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。
谁开发了 Debug Assistant？
由 EvolinkAI（@evolinkai）开发并维护，当前版本 v1.0.0。
推荐 Skills