CONTEXT_OSALPHA

Allye overview

Find the right way to use Allye for your team, your AI agent, or both.

What is Allye?

Allye is a workspace for organizing team work and the context that helps people and AI agents do that work well. It brings delivery tracking, shared documentation, durable memories, and reusable skills into one place.

Use Allye when you need to turn an intention into trackable work, keep a team aligned on how work should flow, or give an agent the context it needs to act without relying on a single conversation.

What you can find here

Choose the result you need rather than starting with a feature list:

If you want to…Start with…
Plan and track workWork items, with boards and workflow for day-to-day execution
Keep delivery organizedBoards, workflow, and sprints
Preserve decisions and project contextMemories
Create knowledge people and agents can return toDocumentation
Give agents repeatable instructionsSkills
Control who can access and change workspace dataAccess and teams and Settings and permissions

Agents can use the same workspace concepts to recover context, inspect or update work, navigate documentation, and follow approved skills. The MCP overview explains that agent-facing workflow.

Choose your path

Use the web app

Choose the web app if you want to explore Allye visually, manage team settings, create work items, configure boards, or maintain documentation and memories directly.

Start with Access and teams to sign in, select the active team, and understand permissions. Then create a work item or open an existing board.

Connect an AI agent through MCP

Choose MCP if your agent already supports the Model Context Protocol and you want it to work with Allye as a structured workspace. An agent can search durable context, manage work items, inspect boards and sprints, write documentation, and use shared skills.

Read MCP installation and setup, then review MCP capabilities for the connection model and agent-oriented concepts.

Install the Allye Plugin

Choose the plugin if you want Allye's workflow guidance and skills integrated into a supported AI coding agent. The plugin adds an adaptive workflow, context and continuity practices, verification guidance, and platform-specific skill delivery on top of the agent you already use.

Start with Install the Allye Plugin. Then read Skills and availability and Adaptive workflow to understand how guidance is selected and applied.

Concepts to know

Teams and active context

A workspace can contain multiple teams. The active team determines the default context for team-scoped work, boards, documentation, memories, and permissions. People and agents should confirm the team before creating or changing shared data; a multi-team agent should select a team explicitly rather than guessing.

Work items and hierarchy

Work items represent work that can be prioritized, assigned, placed on a board, and followed through a workflow. Epics, Features, Stories, Tasks, Bugs, and other supported types can form a hierarchy so broad outcomes and execution details stay connected. Status names and allowed transitions are configured by each team.

Boards, workflow, and sprints

A board is an operational view of work. Its columns map to the team's workflow statuses. Kanban supports continuous flow; Sprint methodology organizes work into time-boxed iterations. Moving an item on a board changes its workflow status, so agents should inspect the current board and status mapping before moving work.

Documentation

Documentation is a navigable knowledge tree of pages and folders. It supports Markdown, metadata, search, publication states, and team-based access. Use it for explanations, procedures, and reference material that people and agents need to find again.

Memories

Memories preserve decisions, discoveries, constraints, preferences, and other context beyond the current conversation. They can be searched semantically and connected through relationships. Use a work item for traceable work, Documentation for durable reference material, and Memories for context that should remain recoverable across sessions.

Skills

Skills are reusable instructions for a kind of work. They can define context, procedure, constraints, and the expected result. Agents should check a skill's scope and lifecycle, read it fully, and combine it with the active team, permissions, and current request rather than treating every draft as an approved rule.

Get your first useful result

For a person

  1. Sign in and confirm the active team in Access and teams.
  2. Open Work items and create one item with a clear, outcome-focused title.
  3. Open Boards, workflow, and sprints to place the item in the team's working flow.
  4. Record an important decision in Memories so the context is available later.

For an agent

  1. Connect the agent using MCP installation and setup, or install the Allye Plugin.
  2. Load Allye context and identify the intended team before making a team-scoped change.
  3. Ask the agent to find a relevant work item or decision, and have it report the source context before acting.
  4. Make one small, visible change: update a work item, add documentation, or save a durable decision.
  5. Verify the result in Allye and keep sensitive credentials out of prompts, memories, documentation, and skills.

Next steps