BizzyBees Content Aggregator

BizzyBees Content Aggregator helps you turn multiple web sources into useful, original content — all inside WordPress. It can scan sources you configure (like your own sites or websites that explicitly permit reuse), aggregate the relevant text, and use AI to produce clear, on‑brand summaries, roundups, or refreshes for your posts and pages. You choose how the output is used: create new posts or intelligently update existing content.

Important: BizzyBees uses external APIs (OpenAI; ScrapingBee). See the External services section for what data is sent and links to each provider’s terms and privacy policy. BizzyBees is built with safety, transparency, and WordPress‑native workflows in mind:

• Authorized Sources Only: Intended for content you own, have permission to use, or that explicitly allows republication.
• AI‑Enhanced: Uses OpenAI models (e.g., GPT‑4o and related models) with your API key—so you control quality and costs.
• Automatic Attribution: Every generated piece includes a visible link to the original source.
• Create or Update: Create brand‑new posts/pages or update existing content using Append, Replace, or Merge.
• Scheduling & Background Jobs: Runs automatically via WP‑Cron + Action Scheduler, with detailed logs.
• Emergency Stop: One click pauses everything safely and auto‑recovers.
• Premium & Companion Add‑ons (sold separately): Add Multi‑Task Manager, AI per URL summaries, email notifications and task history.

Who BizzyBees Is For

• Publishers & bloggers who curate topic roundups or refresh evergreen posts.
• Affiliate & review sites that must keep product info current.
• Local/community sites that collect updates from venues, organizations, or official sources.
• Knowledge hubs that summarize and organize authorized domain material for readers.

Highlights At A Glance

• Multiple Sources: Authorized web pages and sites, with optional “scan beyond base URL” using AI link selection.
• AI‑Enhanced Output: Provide prompts that define tone, structure, focus, and title generation.
• Update Modes: Append, Replace, or Merge (full AI rewrite or merge between HTML comment markers).
• Attribution Built‑In: Adds clear source references; optional footnote suppression when transforming your own content.
• Performance‑Conscious: Uses non‑autoloaded options and locking for large aggregations.
• Resilience: Emergency Stop, retry/backoff logic, and FastCGI‑friendly keep‑alive helpers.
• Premium & Companion Add‑ons: Multi‑Task Manager, per‑task API keys, AI per URL summaries, email notifications, task history.

How It Works (4 Steps)

1) Configure Sources Add URLs you’re allowed to use. Optionally enable scan beyond base URL (AI chooses the most relevant internal links to follow) and set a max links per page. BizzyBees deduplicates links to avoid re‑processing equivalent pages.

2) Choose Create or Update Pick whether to create a new post/page or update an existing one. If updating, choose Append, Replace, or Merge. Merge can be “full AI rewrite” or “between markers” using custom HTML comment boundaries for precise insertions.

3) Set The AI Prompts & Rules Give clear instructions for style, structure, voice, and emphasis. You can also define a separate title prompt. Optionally guide HTML structure (e.g., headings, lists, callouts). If you’re transforming your own content and don’t want footnotes, you can disable them.

4) Schedule & Monitor Pick a schedule (15 minutes → monthly) or run on demand. BizzyBees processes in the background via WP‑Cron + Action Scheduler. Use the Logs page to watch progress, retry behavior, and key events. Use Emergency Stop any time to pause everything gracefully.

Why People Use BizzyBees

• Create new content quickly from multiple authorized inputs.
• Keep evergreen content fresh (prices, specs, availability, new sections).
• Summarize niche topic updates into digestible roundups.
• Reduce manual curation time while maintaining editorial control.
• Maintain attribution practices that respect original creators.

Use Cases (with Practical “Mini‑Playbooks”)

These examples are illustrative. Always ensure you have the rights to aggregate and transform the sources you use.

1) Tech Review Sites — Automated Product Refresh

Goal: Keep product specs, known issues, and pricing notes current across dozens of posts.

Suggested Setup: – Sources: Manufacturer product pages, official changelogs, authorized retailer pages, official news/update pages. – Mode: Update Existing posts with AI‑Merge (full rewrite or between markers for the “Specs” section). – Schedule: Hourly or daily depending on product cadence. – Prompt ideas: “Summarize only spec changes vs previous known values; highlight firmware updates; keep tone neutral and factual; no marketing superlatives.”

Tips: – Use between markers (<!--SpecsStart--> ... <!--SpecsEnd-->) so the plugin only adjusts that section. – Keep the rest of your editorial content intact. – Companion AI per URL add‑on can summarize or focus on a particular product on a specific page separately, then your main task can weave that summary into larger comparisons.

2) Travel Blogs — Destination Guide Refresh

Goal: Keep attractions, transport, and local advisories current.

Suggested Setup: – Sources: City or tourism board pages, official transit websites, authorized local listings. – Mode: Update Existing + Append new “What’s New” sections monthly. – Schedule: Weekly or monthly depending on location volatility. – Prompt ideas: “Summarize operational changes (open/closed, time changes), significant new attractions, and transport updates; add a short traveler’s tip list; keep to 300–700 words.”

Tips: – Use AI link selection for scan beyond base URL when a city page links to multiple subsections. – Set a reasonable max links per page (e.g., 3–5) to focus on the most relevant updates.

3) News Roundups — Thematic Summaries

Goal: Compile perspectives from several trusted outlets on a developing theme.

Suggested Setup: – Sources: Topic hubs or dedicated news pages that permit aggregation; your own newsletters or bulletins; official statements. – Mode: Create New posts (daily roundup) + Update Existing “developing story” pages via Append. – Schedule: Daily or twice daily. – Prompt ideas: “Produce a balanced, multi‑angle summary; avoid duplicating large quotes; include a short ‘What Changed Today’ section and a timestamp.”

Tips: – Attribution is non‑negotiable here—ensure sources are always cited clearly. – Use the Logs page to verify which links were followed and when. – Use AI per URL (companion add‑on) to create concise per‑article notes that can be embedded into a broader roundup.

4) Recipe Collections — Seasonal content

Goal: Highlight seasonal themes and keep nutritional notes updated.

Suggested Setup: – Sources: Your own recipe posts; authorized nutrition standards or public domain references. – Mode: Create New collection posts; Update nutrition footnotes via between markers. – Schedule: Seasonal (quarterly) or monthly mini‑refreshes. – Prompt ideas: “Generate a seasonal roundup with headings by course; for each recipe include 1–2 fresh tips; keep total under 1,200 words.”

Tips: – If you’re summarizing your own recipes, you may disable footnotes. – Keep each collection tight and purposeful; overlong lists dilute value.

5) SaaS Comparison Sites — Feature & Pricing Changes

Goal: Watch for changes in pricing tiers and feature checklists.

Suggested Setup: – Sources: Official pricing pages, release notes, changelogs (authorized). – Mode: Update Existing comparison pages using AI‑Merge (between markers for the comparison table). – Schedule: Weekly or daily based on sector volatility. – Prompt ideas: “If pricing changed more than 10%, call it out; update features table; ensure table remains consistent with our columns; keep commentary neutral.”

Tips: – Use HTML comment markers to limit AI changes strictly to the table region. – Consider per‑task prompts (Premium add‑ons) if you track several product categories differently.

6) Local Events Hubs — Automated Calendar & Roundups

Goal: Maintain a rolling list of upcoming events from multiple venues.

Suggested Setup: – Sources: Venue/event pages that allow reuse; authorized listings; local organizations. – Mode: Create New weekly roundups; Update Existing “This Month” page by Replace or Merge with section markers. – Schedule: Daily or weekly. – Prompt ideas: “Summarize events with date/time/location, one‑line hook, and link to official page; group by weekend; include a short ‘Editor’s Picks’ section.”

Tips: – Always verify permissions and attribution. – Keep schedules lightweight; large, frequent crawls may increase API costs.

Detailed Walkthrough

A) Sources & Link Following

• URLs: Add each source you’re allowed to use. Update and archive pages work well for high‑frequency content; evergreen landing pages are better for static sections.
• Scan Beyond Base URL (Optional): When enabled, BizzyBees uses AI to classify and follow a small number of relevant links from a given page. Configure max links per page (e.g., 3–5).
• Deduplication: BizzyBees avoids re‑processing the same canonical page, reducing redundant API usage.
• Per‑URL Summaries (Companion Add‑on): If you install the AI per URL add‑on, you can flag specific URLs for individual AI summaries. Those summaries are inserted inline (wrapped in HTML markers) while the main task still produces an overall post.

B) AI Configuration

• Content Prompt: Be explicit (tone, structure, target audience, length). Example: “Write a concise roundup with H2 headings, bullet points, and a 2‑sentence intro; emphasize practical changes; avoid fluff.”
• Title Prompt: Separate, so you can e.g., “Generate 5 candidate titles, output the strongest one only.”
• Formatting Guidance: You can allow headings, lists, tables, etc. If you want very clean output, instruct the model accordingly.
• Attribution Footnotes: On by default; you may disable them if the source is your own content or attribution is otherwise not needed.

C) Publishing Modes

• Create New: Choose post type and status (draft/publish).
• Update Existing:
  • Append: Adds new material at the end.
  • Replace: Overwrites entire content.
  • Merge: Uses AI rewrite, combining your old content with the new content or operating between markers (<!--Start--> … <!--End-->) for controlled edits.

D) Scheduling & Background Jobs

• Intervals: Every 15 minutes, hourly, twice‑daily, daily, weekly, monthly, or manual only.
• Processing: Runs via WP‑Cron and uses Action Scheduler for background queues. This keeps admin screens responsive and spreads work safely over time.
• Nudging: BizzyBees can “nudge” cron if your host’s WP‑Cron needs a little help staying active.

E) Monitoring & Logs

• Logs Page: See step‑by‑step status (fetch attempts, AI status, aggregator length, link choices).
• Key Event Filtering: Prefer concise logs in production; enable verbose during setup.
• Emergency Stop: Instantly sets a stop flag, cancels in‑flight work, and clears transient queues. Auto‑clears after 5 minutes (or clear manually).

Performance & Scalability

• Internal Buffering: Aggregated content is stored using non‑autoloaded options, with size checks and simple locking, so large content doesn’t bloat your options table or page loads.
• Non‑Autoload Options: Heavy data is saved without autoload to keep page loads fast.
• Automatic Trimming: BizzyBees trims input to 450,000 characters, to prevent astronomically large prompts. If retries fail more than four times then the token limit is progressively reduced.
• Keep‑Alive & Timeouts: Designed to cooperate with slower hosts such as FastCGI/mod_fcgid setups; includes retry logic and exponential backoff for transient issues.
• Concurrent Work: The Premium Multi‑Task add‑on isolates tasks so one backlog doesn’t block another. Use conservative schedules if your host is resource‑constrained.

Cost Planning Because you connect your own OpenAI and ScrapingBee accounts, you control spend. Practical tips: – Start with conservative schedules and shorter prompts. – Limit “scan beyond base URL” to 3–5 links per page. – Use deduplication (on by default) to avoid re‑scanning similar pages. – Set token caps where appropriate.

Premium & Companion Add‑ons (Sold Separately; not required for the free plugin)

What they add: – Multi‑Task Manager: Create unlimited independent tasks with their own URLs, prompts, schedules, and publishing targets. – Per‑Task API Keys: Use different OpenAI/ScrapingBee keys per task. – AI per URL Summaries: Flag specific URLs for individual AI processing, with optional per‑URL prompts and inline HTML markers. – Email Notifications: Per‑task success/failure summaries and content previews via email. – Task History: Quick previews of the last outputs per task.

The free plugin functions without any add‑ons. These extras are distributed outside WordPress.org.

Third‑Party Libraries

Action Scheduler We bundle the official Action Scheduler library (used by WooCommerce and others), included unmodified for maximum compatibility. If automated code scanners flag direct SQL calls or a different text domain inside lib/action-scheduler/, those belong to Action Scheduler’s own codebase, not BizzyBees custom code.

External services

BizzyBees connects to third‑party services to perform its core functionality. These connections are only made when you provide the relevant API key(s) in the plugin settings and run a task.

OpenAI API (required)

• Purpose: Generate summaries and merged updates from the scraped page text and your task instructions.
• Data sent: Your prompt/instructions, the text extracted from the URLs you configured, and task metadata required to format the result. Your OpenAI API key is sent in the request headers to authenticate the API call.
• Provider policies:
  • Terms: https://openai.com/policies/may-2025-business-terms
  • Privacy: https://openai.com/policies/row-privacy-policy

ScrapingBee API (required)

• Purpose: Fetch HTML/text from pages.
• Data sent: The target URL(s) you configure and ScrapingBee request parameters (e.g., country), plus your ScrapingBee API key for authentication.
• Provider policies:
  • Terms: https://www.scrapingbee.com/terms-and-conditions/
  • Privacy: https://www.scrapingbee.com/privacy-policy/

Consent & control

• No external calls occur until you enter valid API keys.
• Only the content/URLs you explicitly configure are sent to those services.
• You control when tasks run (manual or scheduled).
• You can disable or remove API keys at any time to stop outbound requests.

Data handling

• BizzyBees does not collect personal data from your site’s visitors.
• The free plugin does not call BizzyBees servers for license checks or analytics (“no phone‑home”).
• Premium/add‑ons are distributed separately outside WordPress.org.
• If an external API is unreachable or returns an error, BizzyBees logs the event and retries safely; it won’t break your site.

User responsibility

• Always confirm you have rights to access, aggregate, transform, and republish content.
• Follow third‑party service terms and applicable copyright/data‑protection laws.

Troubleshooting

I don’t see anything running on schedule. – Some hosts disable WP‑Cron for performance reasons. If so, enable a real cron job to hit wp-cron.php periodically. – Visit the Logs page to confirm whether jobs are queued or blocked by an Emergency Stop flag.

The logs say “content too short” or retries keep happening. – Decrease the min length requirement or increase max links per page if you’re using scan‑beyond. – Ensure your prompts aren’t overly restrictive (e.g., “only include X if Y” may filter too aggressively).

I’m seeing occasional timeouts. – Use more conservative schedules (e.g., twice daily vs hourly). – Choose fewer max links per page or fewer URLs. – Try changing between advanced and normal scanning mode. – If you’re using the AI per URL companion add‑on, use it to keep each page’s extracted content tighter and reduce the final prompt size. – Use a smaller model such as GPT‑4o rather than GPT‑5.

BizzyBees is looping or not working properly – Press the Reset BizzyBees button to clear all transients, locks, queues and schedules. – Avoid running the same task too soon after a previous run has finished; try to leave at least 5 minutes between runs.

Attribution looks wrong or is missing. – Double‑check that sources are accessible and the text was actually included in the aggregation. – If you disabled footnotes for your own content, ensure you …