Overview
The Brain Explorer is an interactive dashboard for browsing all the memory items stored in your project's brain. It shows categories, item details, freshness metrics, and an interactive knowledge graph.
Navigate to Brain Explorer from the sidebar.
Stats panel
The top of the page shows four key metrics:
| Metric | Description |
|---|---|
| Total Items | Number of memory items (architecture, decisions, bugs, etc.) in the brain |
| Categories | Top-level categories organizing your items |
| Tokens | Estimated token cost when loading the full brain into AI context |
| Schema Keys | Unique schema keys used to organize and reference items |
Memory freshness
The freshness panel shows how up-to-date your brain is:
- Freshness score — Percentage bar showing overall brain freshness
- Fresh — Items updated within the last 7 days
- At Risk — Items not updated for 7–30 days
- Stale — Items not updated for over 30 days
Click Audit to run a freshness check that flags outdated entries. If stale items are found, a warning banner suggests running a Genesis scan to refresh.
Category distribution
A bubble chart shows the relative size of each category. Larger bubbles indicate more items. The brain organizes knowledge into 27+ categories spanning the full schema key taxonomy: Architecture, Conventions, Decisions, Product, Stack, Auth, Security, Data, Performance, Infrastructure, Testing, Docs, Frontend, Backend, Data Flow, Entry Points, Dependencies, CLI, MCP, VS Code, Journal, Bugs, Todo, Code Map, and more. Categories from both Genesis Scan and session enrichment are unified into a single view.
View modes
Toggle between two views using the switcher in the top-right corner:
Tree view
The default view. A two-panel layout:
- Left panel — Hierarchical tree of categories and sub-categories, searchable by name or schema key. Click a node to load its items.
- Right panel — Item detail list for the selected category.
Item detail panel
The detail panel shows all items under the selected schema key with:
- Status tabs — Filter between Active, Archived, and Review items
- Search bar — Filter items by title, content, or tags
- Type filter — Filter by item type (Decision, Convention, BugFix, Todo, ArchitectureNote, ImplementationFact, CodeMapNode)
- Confidence filter — High (80%+), Medium (50–79%), Low (below 50%)
- Sort — By most recent, confidence, or alphabetical
Each item card shows:
- Title, type badge, and tags
- Importance stars (1–5) and confidence score
- Staleness indicator (age in days for items older than 30 days)
- Expandable details: full facts, rationale, impact, file paths, and code snippets
- Actions — Edit (inline editor), Archive/Unarchive, and feedback buttons (thumbs up/down to adjust confidence)
Graph view
An interactive force-directed knowledge graph powered by react-force-graph-2d. Nodes represent categories and sub-categories, with the central "Brain" node connecting everything. Node size reflects item count. Click any node to jump to its items in tree view.
A color-coded legend appears at the bottom of the graph.
Copy brain
Click Copy Brain in the top-right to copy the entire brain document to your clipboard. This is useful for pasting into AI tools that don't have MCP integration.
Prerequisites
The Brain Explorer requires memory items in your project. Populate the brain by:
- Running a Genesis Scan to analyze your codebase
- Saving session data via the CLI or MCP server
Next steps
- Memory — Manage memory with sessions, review queue, and hygiene
- Genesis Scan — Run a codebase analysis to populate the brain
- Findings Explorer — Browse Genesis findings across all layers