WP Manifestindependent plugin directory
manifest / ai / chatbot_lite

Dual Chatbot Plugin

Provides a dual chatbot for FAQ and member advisory within WordPress. Implements server‑side RAG retrieval and integration with OpenAI for conversational support.

by Nemanja Stojakovic · github.com/hermetik1/chatbot_lite · website

★ 0stars

Install

Download source zip No release zip yet

This is the repository archive. It installs, but the folder name will carry the branch suffix and updates will not flow.

Install straight from GitHub with WP-CLI:

wp plugin install https://github.com/hermetik1/chatbot_lite/archive/refs/heads/master.zip
No built-in updater. New versions won't appear under Dashboard → Updates. Install new releases yourself, or let Git Updater track the repository.
Readme github.com/hermetik1/chatbot_lite ↗

WordPress plugin with dual chatbot system (FAQ + member advisory)

Updates

  • Add .editorconfig for consistent formatting (PHP 4-space, JS/CSS 2-space).
  • Extended error logging:
    • Enabling “Debug-Log” writes to wp-content/uploads/chatbot-debug.log.
    • Client errors (window errors/unhandled rejections) are captured and sent to REST dual-chatbot/v1/client-log when debug is enabled.
  • Improved input validation:
    • New setting “Max message length” (default 2000 chars).
    • Client trims/clamps; server validates and rejects overlong inputs.

Build System

  • Uses esbuild to produce assets/js/chatbot.min.js and assets/css/chatbot.min.css.
  • Commands: npm install, then npm run build.

API Versioning (Prep)

  • REST namespaces are filterable: defaults chatbot/v1 and dual-chatbot/v1.
  • Filters: dual_chatbot_api_version, dual_chatbot_api_ns_public, dual_chatbot_api_ns_internal.

Multi‑Server Support

  • Rate limiting uses shared object cache when available.
  • X-Request-Id header for correlation in all responses (including streaming meta).
  • Logging destination customizable via filters (error_log or file path).

Performance Monitoring

  • Server: inserts perf_server events (latency_ms) into analytics table.
  • Client: sends perf_ui (widget build time) to analytics endpoint.

Note: Enable the debug log only temporarily in production.