Roadmap

IntelliJ IDEA and JetBrains IDEs native integration

New memory tools for Claude and other MCP-compatible agents

Full migration to hashed key storage for all integrations

Auto-capture PRs, reviews and commit context as memory

Receive memory digests and session summaries in Slack

Cross-member activity insights and knowledge coverage reports

Subscribe to brain updates via outbound webhooks

Bring your own AI model for enrichment and Ask AI

Stable versioned REST API with OpenAPI spec and SDKs

Detect convention violations across sessions

Auto-generated new developer guide

Live service health monitoring

Change impact analysis across modules

Auto-generated changelogs from sessions

Light theme for docs and landing

Responsive mobile experience

Cosine similarity at 0.88 threshold

Merge genesis + session overlaps

Post-synthesis final review by Gemini

Regex patterns + short description detection

Per-layer exclusion rules in prompts

Auto-stale detection + review flagging

Real-time 8-phase scan progress

5-step guided setup with Genesis scan

Scan history and quality trends

Token-budgeted memory (~2K tokens)

Genesis → canonical schema keys

Product, stack, auth, data, perf, infra...

All enrichment keys have brain sections

Vulns separated from technical debt

Scan / Session / Manual on every item

Relative timestamps on all findings

27 tabs, source badges, heatmap

A–F scoring, severity, auto-dismiss

Q&A over memory with citations

Force graph, tree view, bubbles

On-Call, Review, Problems, Hygiene

Files, events, timeline per session

Product → architecture → security → risks

OWASP vulnerability detection + scoring

Codebase structure visualization

Technical debt and config concerns

5-pass deduplication pipeline

AI confidence 0-100% per finding

Embedding-powered search

Token-budgeted relevant context

Fast loading with HTTP ETags

Q&A with citation enforcement

Semantic search with category filter

Focused markdown context builder

Analyze then apply changes

Stale + low-confidence triage

Git commits → archive resolved items

Reject hallucinated citations

Small/Medium model by plan tier

Optimized 3K chars/commit diffs

Structured knowledge from raw events

3-layer tiered knowledge document

30 keys with auto-classification

SHA-based range tracking + patches

Project → 15-20 sub-contexts

Directed links: depends-on, see-also

Poll HEAD every 5s, capture diffs

Before/after file modifications

Auto-create, reuse, close sessions

Signed endpoint with replay protection

CLI background auto-capture

Ingest from external sources

Full memory CRUD, search, hygiene, scan

One-click setup for all AI tools

Memory, save, scan, push, pull, hygiene...

cursorrules, copilot, markdown

Context tree with session watcher

Pushes relevant memory to AI tools

Connect from dashboard in one click

Native MCP with full tool access

MCP or CLI with .cursorrules

MCP or CLI with auto-context

CLI with copilot instructions

MCP-compatible for any AI tool

Event hooks for external systems

HMAC signing, rate limits, error codes

Email/password, verification, reset

Scoped keys for programmatic access

Invites, roles, seat management

Stripe plans, credits, invoices

Admin-only user & project management

125+ MDX pages with Cmd+K search

Dark mode SaaS with animations

MDX-powered with tag filtering

Hierarchical CRUD and search