OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
Agent Memory & RAG

Powercontext

Powercontext is a Python library and CLI for preserving task context as people and AI agents hand work off and continue it across conversations. It organizes durable information as Memory, current objectives and state as Handoffs, and reusable approaches as Experience or Skill.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Powercontext — OpenViST project profile in Agent Memory & RAG.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use Powercontext when work is likely to move between people or agent sessions and the next participant needs a structured account of what has happened and what remains.

Who it is for

  • Teams handing partially completed work between human operators and agents.
  • Agent builders who need context to persist across conversations or sessions.
  • Developers prototyping context-preserving handoffs in multi-agent workflows.

Documented uses

  • Carry decisions, constraints, progress, evidence and next steps into a later session.
  • Let a new person or agent continue a task without rereading the full conversation history.
  • Keep durable memory and the current work state organized within the task's scope.
  • Record reusable approaches as Experience or Skill.
Before you choose

Requirements & limits

The documented installation uses a CLI and Server. The project lists macOS and Linux support and describes Windows support as experimental.

  • Basic memory supports explicit saving and retrieval without additional model APIs; full memory capabilities require separate Generation and Embedding API credentials.
  • The project identifies Codex integration as official, other hosts and Python agent frameworks as community, and Bub as evaluation only.
  • The setup requires Git, uv and Python 3.11 or later; uv can provision Python.
  • The repository identifies the license as Apache-2.0.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Set up PowerContext 1.2.0 and connect Codex

Install and configure PowerContext, start its local Server, and install the matching Codex integration.

Before you start

  • macOS or Linux; Windows is experimental.
  • Python 3.11+, Git, uv, and an installed Codex CLI.
  • For full memory, working Generation and Embedding model APIs with their base URLs, model names, and API keys. Without these, choose basic memory for explicit saves and recall.

Steps

  1. Install PowerContext and create a dedicated configuration directory: the command below, then mkdir -p ~/powercontext-demo and cd ~/powercontext-demo.

    Documented command
    uv tool install --force "powercontext[cli,server]==1.2.0"

  2. Run powercontext config init --language en --output .env. Choose storage, local or remote access, memory capabilities, Dashboard, and Codex integration in the wizard; review and save its .env and .env.next-steps.md files. Do not commit these credential-containing files.

  3. Validate the configuration and start the Server with powercontext config validate --env-file .env and powercontext server run --env-file .env. Keep the Server terminal running.

  4. In another terminal, load the client settings and check the running service: cd ~/powercontext-demo, set -a, . ./.env, set +a, then run powercontext ready and powercontext capabilities.

  5. Follow .env.next-steps.md to create the planned Scope and use the real returned scope ID in .env; reload the settings and run powercontext setup codex --ref powercontext-v1.2.0 and powercontext doctor codex.

Expected outcome

PowerContext is configured, the Server readiness and selected capabilities can be checked, and the matching Codex plugin setup can be verified.

Guide limitations

  • Setup and connectivity checks do not verify memory capture, topic evolution, or cross-session recall; the quick-start walkthrough describes those subsequent checks.
  • Full memory requires separate Generation and Embedding API credentials. Basic memory supports explicit saves and recall but does not enable automatic Topic Memory.
  • The Codex plugin installation does not start the Server. Keep the Server running, and use the same Scope in the Dashboard and Agent for subsequent checks.
  • The .env files contain credentials and should not be committed.
Documentation 1 ↗Documentation 2 ↗
Evidence and transparency

Sources & review

Official repository ↗Guide documentation ↗
Source record details
Source availability
Active
Sources checked
11 October 2026
Original website title
oceanbase/powercontext · Official homepage title · retained in its source language
How we review sourcesSuggest a correction