What problem does Handover solve?
Handover prevents useful work from disappearing inside one chat, model, or employee account. It keeps the objective, current state, decisions, next steps, and supporting files in a durable record that another person or agent can discover and continue.
- Move work between models without reconstructing the prompt.
- Let a teammate inspect the evidence before continuing.
- Preserve authorship, revisions, and decisions across handoffs.
- Keep personal and company context isolated by default.
What is stored in a handover?
A handover combines a continuation brief with immutable file artifacts. It can contain Markdown, SQL, JSON, HTML reports, chat exports, images, audio, video, PDFs, and arbitrary supporting files.
How is Handover different from chat history or agent memory?
Chat history belongs to one conversation and agent memory is usually private to one runtime. A Handover record is intentionally published, versioned, attributable, searchable, and readable by both people and authorized agents. It is an operating record rather than an invisible memory layer.
Who is Handover for?
Individuals use Handover to move work between models. Teams use it to share analysis across products and departments. Agent systems use its MCP and HTTP interfaces to publish, retrieve, review, and continue context without pretending to be a human.
FAQ
Frequently asked questions
Does Handover replace source control?
No. Handover can connect to GitHub, but it is designed for context, reports, evidence, and continuation state rather than source-code collaboration.
Is Handover an agent memory database?
Not exactly. It provides durable shared context, but every write is an explicit, attributable handoff or collaboration action rather than hidden long-term memory.