Project Home: the continuity hub for long projects in Navin

Coming back to a project after three weeks and being productive within a minute: that is exactly what Project Home, Navin's continuity hub, is for. One screen gathers the resume brief, the task board, GitHub issues, the Vision 360 audit, the dependency graph, the project brain and the git timeline - and a Resume button relaunches the agent right where everything stopped.

The problem nobody solves well

AI IDEs are excellent within one session. But a real project lives for weeks: decisions made, constraints discovered, half-finished tasks, open issues. When you come back, all that context is scattered across your memory, some notes and chat history. Navin structures it inside the repository itself: the .navin/ folder travels with the project, like the code.

One screen, eight synchronized tabs

The detail that changes everything: Tasks, Vision 360 and Graph are not copies. They are the same components as the Code workbench, wired to the same APIs and the same live events. A card moved here is moved everywhere, including for the agent.

The project brain travels with the repository

Every project automatically receives its continuity pack, without overwriting anything that exists:

• RESUME.md: the resume brief, updated automatically at every session compaction.
• DECISIONS.md: long-term decisions, so a settled choice is never re-debated.
• memory/MEMORY.md: the project's durable facts.
• SOUL.md and USER.md: the agent's personality and your preferences, scoped to the project.

These files are injected into the agent's context on every turn. As a result, the studios and the Code module share the same brain, and an imported project (not created from scratch) gets the same treatment, right next to its existing .cursor or .claude folders.

One-click resume

The Resume tab assembles a resume seed: the RESUME.md brief, the constraints from MEMORY.md and the top open tasks. The Resume button opens a chat pre-filled with that seed: the agent knows where the project stood, what is blocked, what is ready, and proposes the next move - even after weeks away.

Recommended workflow for a long project

1. Open Project Home, read the Resume brief (1 minute).
2. Check the Tasks tab: ready tasks, blocked tasks, critical path.
3. Enable board autonomy if the project calls for it.
4. Click Resume and let the agent chain through the ready tasks.
5. Review the opened PRs, merge, and let the timeline document the rest.

FAQ

Does Project Home replace the Code workbench?

No, it complements it: Code is the workstation of one session, Project Home is the long-run view of the project. Both share the board, the audit and the graph.

Does the .navin/ pack overwrite my existing files?

Never. Creation is idempotent: what exists is preserved, only what is missing is added.

Do the studios see the same brain?

Yes. Memory, decisions and the board are scoped to the project, not to the chat. A marketing studio and the Code module read the same files.

What if the project has no GitHub remote?

The Issues tab says so clearly, and everything else (Resume, Tasks, Brain, Graph, Vision 360) works without it.

Download Navin · Features · Documentation

Recommended next reads

• The autonomy that executes your tasks → autonomous board
• Plan / enrich / steer (human, agent, hybrid) → expert board
• Plan / Mission (/blueprint, /forge, /cruise) → Plan Mission
• Durable memory → Project Brain
• Git kill-switches → Git kill-switches
• The control room of one session → Vision 360 workbench
• The 360° health audit / PDF → Vision 360 audit
• The full coding cycle → Code module
• An agent team on the project → Navin agents

Conclusion

Continuity is not a chat feature, it is an architecture: a versioned project brain, a board shared by humans and agents, and a one-click resume. That is what makes Navin a tool for long projects, not just for brilliant sessions.