內容簡介
Zybroz AI Agents 是一款專為 WordPress 設計的外掛，幫助使用者創建、管理和運行 AI 驅動的代理。透過結構化的 JSON 配置，使用者可以自動化各種 WordPress 任務，提升網站管理效率。
【主要功能】
• 視覺代理建構器：直觀的兩欄式介面，實時 JSON 預覽
• AI 提示生成器：用自然語言描述需求，自動生成系統提示
• 聊天創建向導：引導式流程，快速創建代理草稿
• 代理版本控制：自動版本更新及歷史記錄管理
• 匯出/匯入功能：支持 JSON 和提示模板格式的匯出與匯入
• 多模型 AI 支援：兼容多種 AI 模型，標示缺失的提供者
外掛標籤
開發者團隊
原文外掛簡介
AI Agents for WordPress helps you create, manage, and run AI-powered agents inside WordPress. Each agent is defined by a structured JSON configuration with persona, scope, logic, safety, and versioning rules, so you can automate WordPress tasks with clear controls and repeatable behavior.
It is designed for site owners, marketers, agencies, developers, and SMBs who want practical automation beyond basic text generation. The plugin currently covers content rewriting, image alt text, internal linking, schema review, comment handling, plugin updates, backups, and other WordPress operations through a mix of AI providers and logic-only actions.
The plugin is built around a WordPress-native workflow:
– Build an agent in the Visual Agent Builder.
– Validate it with Quality Precheck.
– Test it with Dry Run.
– Qualify it after successful execution.
– Use it locally or publish it later through the library flow.
Key Features
Visual Agent Builder — two-column builder with tabbed panels for Persona, Scope, Logic, and Safety, plus live JSON preview. Save and qualify is gated behind a successful dry run.
AI Prompt Generator — describe what you want in plain English; AI writes a structured system prompt; compare, accept, keep, or regenerate.
Chat-to-Create Wizard — guided creation flow that suggests agent name, role, goal, system prompt, and category, then creates a draft and runs a dry run.
Agent Versioning — automatic patch bumps on save, stable and major release bump actions, plus version history and restore.
Export / Import — export agents as JSON or as a Prompt Template text file for use in external AI tools; import either format back into the builder.
Multi-model AI — support for Anthropic Claude and local Ollama, with unavailable models clearly labeled when a provider is missing.
BYOK Security — API keys are encrypted before storage and are never exposed to the browser.
InbuiltEngine — logic-only AI Actions can run with no API key for maintenance and admin tasks.
Agent Library — ready-to-use starter agents across content, SEO, images, schema, social, comments, and admin health.
AI Action Library — vetted atomic operations for real WordPress changes, including read and write actions.
Advanced Backup Module — full DB backups, uploads backups, list/delete/restore, and protected storage in wp-content.
Quality Precheck — validates agent JSON before qualification and highlights missing fields, scope issues, or missing provider settings.
Favorite System — installed agents can be marked as favorites, with metrics updated from favorites instead of votes.
Safety Layer — automatic post snapshots, dry-run preview mode, and one-click rollback before write operations.
Status Machine — agents move through draft, active, and failed states with clear UI indicators.
Run History & Metrics — track total runs, success rate, and favorites for each installed agent.
Dashboard — see site health, recent runs, provider connection status, and quick actions in one place.
Starter Agents
Traffic Healer — detects and rewrites underperforming content.
Alt Text Architect — generates descriptive alt text for images missing descriptions.
Internal Link Weaver — suggests contextual internal links for your posts.
Schema Guardian — audits schema markup and flags missing structured data.
Social Sniper — generates tailored social media posts from existing content.
Requirements
PHP 8.1 or higher.
WordPress 6.4 or higher.
Anthropic API key for Claude models or Ollama running locally.
Privacy
No data is sent to third parties unless you configure an external API key.
Depending on which AI Agents and AI Actions you run, the content sent to your configured AI provider can include WordPress content such as post titles/body, and — for comment-moderation agents — commenter names and comment text. See “External Services” below for exactly what is sent and when.
API keys are stored encrypted.
Ollama runs locally with no external calls.
No telemetry or call-home behavior.
External Services
This plugin can optionally connect to three external AI providers. All connections are user-initiated and require you to supply your own API key in Settings → API Keys. No data is sent to any provider without your key being configured and an agent run being triggered.
The exact content sent depends on which agent or AI Action you run. This always includes the agent’s configured system prompt. Depending on the agent’s scope, it can also include WordPress content relevant to that agent’s task — for example, a post’s title and body for a content-rewriting agent, or a commenter’s name and comment text for a comment-moderation agent. The plugin does not send WordPress credentials, API keys, or payment information to any AI provider. If your site handles content containing personal data (e.g. names in comments or post content), that data may be included in the request when an agent processing that content is run — review each agent’s scope before running it on content containing personal data.
Anthropic Claude API
What it does: Sends your configured agent prompts and receives AI-generated text responses for WordPress automation tasks.
Data sent: The system prompt, plus whatever WordPress content the agent’s scope includes (e.g. post title/content, or commenter name/comment text for comment-moderation agents). No WordPress credentials are sent.
When: Only when an agent run is triggered and Claude is the selected model, or when you click “Test Connection” on the Settings page (calls /v1/models to validate the key only — no WordPress content sent).
Provider: Anthropic PBC, San Francisco, CA.
Terms of Service: https://www.anthropic.com/legal/consumer-terms
Privacy Policy: https://www.anthropic.com/legal/privacy
OpenAI API
What it does: Sends your configured agent prompts and receives AI-generated text responses when an OpenAI model (gpt-4o, gpt-4o-mini, gpt-4-turbo) is selected.
Data sent: The system prompt, plus whatever WordPress content the agent’s scope includes (e.g. post title/content, or commenter name/comment text for comment-moderation agents). No WordPress credentials are sent.
When: Only when an agent run is triggered and an OpenAI model is selected, or when you click “Test Connection” for OpenAI on the Settings page (calls /v1/models only — no WordPress content sent).
Provider: OpenAI, LLC, San Francisco, CA.
Terms of Service: https://openai.com/policies/terms-of-use
Privacy Policy: https://openai.com/policies/privacy-policy
Google Gemini API (Google Generative Language API)
What it does: Sends your configured agent prompts and receives AI-generated text responses when a Gemini model (gemini-2.0-flash, gemini-1.5-pro, gemini-1.5-flash) is selected.
Data sent: The system prompt, plus whatever WordPress content the agent’s scope includes (e.g. post title/content, or commenter name/comment text for comment-moderation agents). No WordPress credentials are sent.
When: Only when an agent run is triggered and a Gemini model is selected, or when you click “Test Connection” for Gemini on the Settings page — no WordPress content sent.
Provider: Google LLC, Mountain View, CA.
Terms of Service: https://ai.google.dev/gemini-api/terms
Privacy Policy: https://policies.google.com/privacy
AI Agents Central Library (Optional)
What it does: Fetches a catalog of community-contributed starter agents from the central registry so you can browse and install them from Agent Library. No agent data or site information is sent.
Data sent: None. Read-only GET requests to the registry endpoint to retrieve the public agent catalog.
When: Only when you explicitly enable “Connect to central library” in Settings → Agent Library. Disabled by default.
Provider: zybroz, https://zybroz.com
Registry endpoint: https://registry.ai-agents.zybroz.com
Terms of Service: https://zybroz.com/terms
Privacy Policy: https://zybroz.com/privacy
