# HeapFile

> HeapFile is a local-first AI memory and coordination layer for builders. It helps preserve project context, decisions, and handoffs across AI sessions.

Desktop is available now. Mobile support is in development. Available capabilities and provider routes depend on the current client, plan, configuration, and authorization.

HeapFile's source code is not public. Public materials describe product behavior without publishing the proprietary implementation. Local-first does not mean every AI task runs offline; review the selected provider and permissions before using sensitive material.

## Official Pages

- [Documentation](https://heapfile.com/docs.html): Start here for public setup guides, connector documentation, and AI-readable references across both official HeapFile domains.
- https://heapfile.com/vscode.html — VS Code extension setup, agent discovery, explicit memory writes and recall, verification, and troubleshooting.
- https://heapfile.com/agent-setup.html — MCP connections across clients, memory instructions, optional VS Code Local hooks, and agent-led setup with verification.

- [HeapFile](https://heapfile.com/): Product overview and current plan information.
- [Connector overview](https://heapfile.com/#connectors): Human-facing explanation of HeapFile's connector model, domain split, and provider boundary.
- [Desktop access](https://heapfile.com/account.html): Sign-in, plan, and current download options. Review account requirements before installing.
- [Trust and proof](https://heapfile.com/governance.html): Public information about context control, provider boundaries, and reviewable work.
- [Support](https://heapfile.com/support.html): Official contact options for access, billing, downloads, and product feedback.
- [Privacy](https://heapfile.com/privacy): Current privacy policy.
- [Terms](https://heapfile.com/terms): Current product and subscription terms.
- [FAQ](https://heapfile.com/faq.html): Answers about memory, desktop access, plans, and handoffs.
- [AI session context](https://heapfile.com/ai-session-context.html): Practical context and handoff guidance.
- [Evaluate AI memory](https://heapfile.com/evaluate-ai-memory.html): An invented-project exercise, not a measured benchmark.
- [Saved handoff guide](https://heapfile.com/context-guide.html): Explicit save-and-recall exercise with synthetic data, setup requirements, and result checks. Not a customer benchmark.
- [Marketing Assist use case](https://heapfile.com/marketing-assist.html): Practical pattern for preserving blockers, owners, evidence, and next actions when one marketing channel is blocked.
- [Feedback](https://heapfile.com/feedback.html): Product feedback and sanitized crash-report contact options.
- [Technical connector catalog](https://heapfile.ai/connectors.html): Canonical connector readiness, authorization, permissions, and data-route reference.
- [Machine-readable connector catalog](https://heapfile.ai/connectors.json): Release-safe public connector manifest.

Provider names identify compatible or planned systems. They do not imply endorsement, sponsorship, partnership, or authorization. The detailed connector source of truth lives on heapfile.ai, not heapfile.com.