features

Context pinning

Most AI tools don't retain a structured understanding of your project — new chat means starting over, restating decisions, rebuilding context you already gave it last week. Context pinning is how MockingBird avoids that: it keeps a running, durable picture of your workspace that persists across sessions and travels with you when you switch models.

How pinning works

When you bookmark a message or moment worth remembering, MockingBird summarizes it down to the essentials and appends it to your workspace context — not the raw message, a condensed version of what mattered.

That context gets injected into every conversation in the workspace going forward, on whichever model you're using. Switch from Claude to Gemini mid-project, and Gemini already knows what was pinned — it isn't starting cold.

Two layers of memory

• Workspace context
  Project-specific — architecture decisions, active tasks, conventions you're following, things pinned from earlier conversations in that workspace.
• Identity context
  You-specific, not tied to one workspace — how you like things explained, recurring preferences, the working patterns that would otherwise get repeated in every new chat.