web-security（网络安全）

web-security

Purpose

This skill enables developers to implement and audit web security measures based on OWASP Top 10 guidelines, including CSP, CORS, XSS/CSRF prevention, authentication patterns, and dependency scanning. It focuses on protecting web applications from common vulnerabilities like injection attacks and unauthorized access.

When to Use

Use this skill during web application development, security audits, or deployments. Apply it when building APIs, handling user authentication, configuring cross-origin requests, or scanning dependencies for known vulnerabilities. Ideal for projects using frameworks like Express.js or React, or when integrating third-party libraries.

Key Capabilities

• OWASP Top 10 Scanning: Detects issues like SQL injection and broken authentication; use built-in checks via openclaw web-security scan --owasp.
• CSP Configuration: Generates Content Security Policy headers; example: set policy with openclaw web-security csp --policy "default-src 'self'"
• CORS Management: Enforces Cross-Origin Resource Sharing; configure with openclaw web-security cors --allow "https://example.com".
• XSS/CSRF Prevention: Provides sanitization functions and token generation; e.g., inject anti-CSRF in code: const token = generateCSRFToken(); res.setHeader('X-CSRF-Token', token);
• Authentication Patterns: Implements JWT or session-based auth; scan for weaknesses with openclaw web-security auth --check.
• Dependency Scanning: Analyzes npm/yarn packages for vulnerabilities; run with openclaw web-security depscan --path./package.json.

Usage Patterns

To accomplish tasks, invoke the skill via OpenClaw's CLI or API. For scanning, provide project paths and flags; for configuration, output directly to code files. Always set environment variables for authentication, e.g., export $OPENCLAW_API_KEY before running commands. Example pattern: Pipe output to a file for integration, like openclaw web-security scan --output report.json. For code snippets, embed generated security code into your app; e.g., add CSP middleware in Express: app.use((req, res, next) => {res.setHeader('Content-Security-Policy', "default-src 'self'"); next();});

Common Commands/API

• CLI Commands: Use openclaw web-security [subcommand] [flags]. For example, scan a project: openclaw web-security scan --project /path/to/app --key $OPENCLAW_API_KEY. API endpoint: POST to /api/web-security/scan with JSON body {"projectPath": "/path/to/app", "apiKey": "$OPENCLAW_API_KEY"}.
• Subcommands:

- scan --owasp --verbose: Runs full OWASP check; outputs vulnerabilities in JSON. - csp --generate --domains example.com: Creates CSP string; e.g., output: "Content-Security-Policy: default-src 'self' https://example.com". - cors --set --origins "http://localhost:3000": Configures CORS in a config file like {"origins": ["http://localhost:3000"], "methods": ["GET", "POST"]}. - auth --pattern jwt: Generates JWT validation code; snippet: const jwt = require('jsonwebtoken'); const verify = token => jwt.verify(token, process.env.JWT_SECRET);. - depscan --format npm: Scans dependencies; e.g., command: openclaw web-security depscan --path package.json --output vulnerabilities.txt.

• API Endpoints: All commands map to /api/web-security/{subcommand}, requiring authentication via header Authorization: Bearer $OPENCLAW_API_KEY. Response format: JSON with keys like {"status": "success", "data": {...}}.

Integration Notes

Integrate by wrapping OpenClaw calls in your build scripts or CI/CD pipelines. For example, in a GitHub Actions workflow, add: run: openclaw web-security scan --project. --key ${{env.OPENCLAW_API_KEY}}. Use config files for persistent settings, e.g., a .openclawrc file with JSON: {"web-security": {"defaultFlags": ["--verbose"], "apiKeyEnv": "OPENCLAW_API_KEY"}}. If combining with other skills, chain outputs; e.g., use web-security scan results as input for a "web-dev" deployment skill. Ensure API keys are stored securely in env vars like $OPENCLAW_API_KEY and never hardcoded.

Error Handling

Handle errors by checking exit codes and response bodies. Common errors: Authentication failure (HTTP 401) if $OPENCLAW_API_KEY is invalid—fix by verifying the key format. Scan failures (e.g., "Project path not found") return code 404; resolve by providing absolute paths. For XSS prevention, if a snippet fails, catch exceptions like: try {sanitizeInput(userInput);} catch (e) {console.error(e.message); // e.g., "Invalid input detected"}. Parse JSON responses for error details, e.g., {"error": "Vulnerability detected", "code": 400}, and retry with corrected flags. Always log errors with timestamps for debugging.

Concrete Usage Examples

1. Scan for OWASP Vulnerabilities: To audit a web app for injection risks, run: openclaw web-security scan --owasp --project /path/to/app. This outputs a JSON report; then, fix issues by adding code like: const safeQuery = db.escape(userInput); db.query(safeQuery);. Expected output: A list of vulnerabilities, e.g., {"injection": ["SQL in login endpoint"]}.
2. Configure CSP for XSS Prevention: To set up CSP in an Express app, use: openclaw web-security csp --generate --policy "script-src 'self'". Integrate the output into your server code: app.use(helmet.contentSecurityPolicy({directives: {scriptSrc: ["'self'"]}}));. This prevents inline scripts, reducing XSS risks.

Graph Relationships

• Related to cluster: "web-dev" (e.g., shares dependencies for web app builds).
• Connected skills: "auth-management" (for advanced auth patterns), "vulnerability-scanning" (for broader security checks).
• Inverse relationships: Depends on "api-tools" for endpoint testing; provides input to "deployment-pipeline" for secure releases.