Open source · local-first

Keep work movingacross sessions.

PowerContext keeps decisions, constraints, evidence, and current progress with the project. When a person or agent takes over, they can verify the state and continue without reconstructing the last conversation.

From installation to your first memory

Start with a Server, configure the capabilities you need, then connect your Agent. The guide covers local use and access from another device.

Follow the complete guide
  1. 1Install PowerContext

    Use Python 3.11 or later, Git, and uv. Install this branch on the machine that will run the Server.

    uv tool install --force "powercontext[cli,server] @ git+https://github.com/oceanbase/powercontext.git@master"
  2. 2Follow the configuration wizard

    Choose your storage, access scenario, memory capabilities, Dashboard, and Agents. It writes configuration files and tailored next steps.

    powercontext config init
  3. 3Start the Server

    Run from the directory containing your generated .env. Open the Dashboard URL and use the token shown by the wizard, if enabled.

    powercontext server run --env-file .env
  4. 4Connect your Agent and verify memory

    Follow .env.next-steps.md to create or select a Scope and connect your Agent. Verify that a real conversation becomes a Source, then check memory processing if enabled.

Guided setup preview. These instructions use the official oceanbase/powercontext master branch. View this source branch

The session ends before the work does.

You make a decision, change the code, and stop before the task is complete. PowerContext keeps the useful context with the project so the next session can continue from the current state.

PowerContext

Current session

  1. Decisions
  2. Constraints
  3. Evidence
  4. Verified progress

Next session

  1. Relevant context
  2. Current objective
  3. Source links
  4. Next action

The work changes hands.The context stays.

A task may start in one agent and continue in another. PowerContext keeps its knowledge, progress, and proven methods with the project so the next agent can pick up from the current state.

Work begins in a connected agent

Context accumulates within the project scope

The work leaves reusable artifacts

Memory

Durable knowledge

Handoff

Active task state

Experience

Reviewed approach

Skill

Exported procedure

A different agent continues the work