# Wagglet > Wagglet is an AI task-handoff workspace for teams: one person writes an exact AI task, and a teammate claims it and runs it on their own agent subscription. Task handoff never shares accounts, credentials, API keys, or provider balances. Each person runs the work under their own Claude Code, Codex, or other agent subscription, and Wagglet records the delivery for human review. Wagglet is operated by Rockbite Games. This file maps the public site only; the signed-in workspace, its APIs, and unpublished writing are deliberately absent. ## Product - [Home](https://wagglet.com/): Hand exact AI tasks to teammates using their own Claude Code, Codex, or other agent subscriptions—without sharing accounts or keys. - [About](https://wagglet.com/about): Wagglet is an AI task-handoff product for teams, operated by Rockbite Games. - [Wagglet for engineering teams](https://wagglet.com/for/engineering-teams): Wagglet keeps human work, agent instructions, execution context, delivery evidence, and review attached to the same task. - [Wagglet for startups](https://wagglet.com/for/startups): Prepare the task once, then route it to the teammate or agent with capacity to run it without sharing accounts, credentials, or allowances. - [Wagglet for game studios](https://wagglet.com/for/studios): Wagglet separates execution from approval so engineers, designers, producers, and QA can each stay responsible for the part they actually know best. - [How it works](https://wagglet.com/how-it-works): See how a team prepares an exact AI task, hands it to a teammate using their own agent subscription, records delivery, and reviews the result. - [Security](https://wagglet.com/security): How Wagglet scopes Task Handoff and MCP credentials, rechecks permissions, protects Private Drafts and attachments, and keeps provider accounts separate. - [Solutions](https://wagglet.com/solutions): Where each AI work handoff question is answered: the category itself, moving work when a teammate is out of agent capacity, carrying a task between Claude Code, Codex and Cursor, how Wagglet sits alongside Linear, and how an agent connects over MCP. - [Shared AI execution capacity](https://wagglet.com/solutions/ai-execution-capacity): When one developer reaches their Claude Code or Codex limit and a teammate still has room, the prepared task moves to that teammate, who runs it on their own subscription. No plan, credit or token allowance is pooled, transferred or extended. - [AI work handoff](https://wagglet.com/solutions/ai-work-handoff): An AI work handoff moves a prepared task to a qualified teammate who runs it on their own agent account. The five stages, who owns each one, and what stays with its owner when the work changes hands. - [Claude Code handoff](https://wagglet.com/solutions/claude-code-codex-cursor-handoff): What a Wagglet task carries when a coding agent task changes hands: the next action, the repository and branch, the constraints, and the delivery contract — and the one field of the five it does not carry. - [Wagglet and Linear](https://wagglet.com/solutions/linear): Compare Linear's issue-to-agent delegation with Wagglet's person-to-person AI work handoff, including the manual alongside workflow and when Linear alone is simpler. - [MCP integration](https://wagglet.com/solutions/mcp): Compare Wagglet's bounded Task Handoff credential with its named MCP Workspace connection for Claude Code, Codex, Cursor, and other MCP clients. ## Documentation - [AI coding agent handoff](https://wagglet.com/ai-coding-agent-handoff): How to hand an in-progress AI coding task to another developer: what context has to move, a free HANDOFF.md template, and when a single file stops being enough. - [Docs](https://wagglet.com/docs): Official Wagglet guides for bounded Task Handoffs, ongoing MCP Workspace access, permissions, delivery, and human review. - [MCP docs](https://wagglet.com/docs/mcp): Connect an agent to one Wagglet Team with a named MCP Connection, current product permissions, explicit actions, revision checks, and revocation. - [Task Handoff docs](https://wagglet.com/docs/task-handoff): How to run one claimed Wagglet task with bounded agent context, human supervision, a separate delivery prompt, and independent review. ## Agent skills - [Task Handoff skill](https://wagglet.com/skill.md): Agent workflow for one claimed Wagglet task: read the bounded task context, stay inside scope, and report the delivery. - [MCP Workspace skill](https://wagglet.com/skills/wagglet-mcp/SKILL.md): Agent workflow for the Wagglet MCP Workspace: connect to one Team, then search, inspect, and change authorized work. ## Blog - [Field notes](https://wagglet.com/blog): Original Wagglet essays on AI task handoffs, human-agent collaboration, team capacity, incentives, and verified delivery. - [Use the AI capacity your company already pays for](https://wagglet.com/blog/use-unused-claude-code-codex-capacity): Use more paid AI capacity by moving prepared work—not accounts or allowances—to authorized teammates, then separate recovered output from real cash savings. - [The Dual Prompt](https://wagglet.com/blog/dual-prompt-human-agent-task-design): A practical pattern for writing one work item with detailed agent context and a concise human runbook for judgment and verification. - [Bridge small skill gaps with AI task handoffs](https://wagglet.com/blog/close-team-skill-gaps-with-ai-task-handoffs): How expert context, agent guidance, human judgment, and acceptance criteria can turn narrow skill gaps into bounded work. - [More people get to contribute](https://wagglet.com/blog/meaningful-work-with-human-ai-pairs): A human view of AI task handoffs: creating bounded, credited ways for capable teammates to contribute and learn. - [Bounties and rewards for AI-assisted work](https://wagglet.com/blog/gamification-ai-work-bounties-rewards): Sometimes: rewards can speed a real queue when they pay only for verified outcomes, but weak metrics can increase activity without improving useful work. - [The complete Wagglet workflow](https://wagglet.com/blog/wagglet-workflow-request-draft-ticket-delivery): Follow a Wagglet work item from rough request through Draft, claim, agent work, delivery, acceptance, and truthful merge state. - [The AI-native mobile game studio operating stack](https://wagglet.com/blog/ai-native-mobile-game-studio-stack): Fast agent-friendly game tools still need an operating workflow that moves prompts, judgment, evidence, and delivery across a studio. - [Wagglet vs Linear](https://wagglet.com/blog/wagglet-vs-linear-comparison): Linear leads on initiatives, roadmap timelines, and formal milestones. Wagglet extends further into employee availability, workload, AI capacity, and delivery. A researched, honest comparison. ## Optional - [Privacy](https://wagglet.com/privacy): How Wagglet collects, uses, shares, protects and deletes personal data, including the Wagglet mobile app and your privacy rights. - [Terms](https://wagglet.com/terms): The current terms for using Wagglet, including beta access, accounts, teams, acceptable use, service limits, and ending use.