bbuddy works like a posting assistant with memory. It does not write from one prompt alone. Before drafting, it looks at the useful context in your workspace: brand profile, Library, Feed, connected platforms, recent posts, previous suggestions, and saved feedback.
The generation loop
- Understand the request. bbuddy decides whether you are asking a question, drafting a post, editing a draft, building an automation, or scheduling existing content.
- Read context. It pulls the relevant brand, Library, Feed, post history, and platform information before writing.
- Choose the target platforms. If you name platforms, bbuddy uses those if they are connected. If you do not name platforms, it uses the connected channels that make sense for the task.
- Draft per platform. One idea can have different copy for Facebook, Instagram, LinkedIn, X, and other channels. Short platforms get shorter variants; longer platforms can get more explanation. Saved per-platform style rules are applied here, so a platform preference should change the actual draft rather than appear as a label.
- Attach media when needed. Instagram, TikTok, YouTube, and Pinterest require media. bbuddy first looks for an existing asset, then generates an image when the request or platform requires it.
- Save a suggestion. Generated posts land in Suggestions so you can review, edit, schedule, save as draft, post, or reject them.
What bbuddy reads
- Brand profile: name, positioning, audience, voice, colors, website facts, and social links.
- Library: durable memory such as tone rules, product facts, examples, quotes, images, imported pages, and saved feedback.
- Feed: fresh articles, products, social posts, RSS items, events, and other source material.
- Recent posts and suggestions: what has already been drafted, scheduled, rejected, or published, so bbuddy can pick fresh angles instead of repeating itself.
- Connected accounts: which platforms can actually receive posts.
How to steer output
You can give direct instructions in chat. The best prompts say what to create, where it should go, and what source or angle to use.
- "Create a Facebook and Instagram post about this new product."
- "Turn the latest feed article about pricing into a LinkedIn post."
- "Make this more founder-led and less corporate."
- "Use a thread on X and explain the point plainly."
- "Save that as a rule for future retention posts."
What feedback does
Feedback can become memory. When you say "use more threads", "avoid teasing", "write in native French", or "retention posts need images", bbuddy can save that as a Library rule. Future generations then read it before drafting.
How bbuddy avoids repetition
bbuddy uses recent posts and suggestions as a map of what has already been covered. When you ask for a new batch, it should look for open ground before writing instead of producing several drafts around the same angle.
If a draft is too close to existing content, ask for a different angle, audience, story, product moment, or source. Specific direction helps more than "try again".
Media logic
- Existing media first: bbuddy prefers images and videos already in your Library or attached to relevant source items.
- Generated images second: if no existing asset fits and the platform or request needs a visual, bbuddy can generate one.
- Instagram rule: Instagram cannot publish text-only posts through the API, so every Instagram suggestion needs at least one image or video.
Why a draft may fail or be blocked
- The selected platform is not connected or was disconnected.
- The draft is too long for a selected platform.
- A media-required platform has no attached media.
- The draft needs a fresher angle than recent posts or suggestions.
- The social platform rejects the publish request.