oceanbase/powercontext

Not only memory but a full story.

What it solves

PowerContext addresses the problem of context loss when tasks are handed off between humans and AI agents. In typical workflows, reasoning and current state are often trapped within a specific conversation history, making it difficult for a new person or agent to resume work without rereading the entire history.

How it works

It acts as a durable context layer that stores and organizes information across conversations. It categorizes information into different types:

  • Memory: Durable information stored for later use.
  • Handoff: The current objective and state of a task.
  • Experience and Skill: Reusable approaches and methods.

Users can integrate PowerContext with various AI agents (such as Claude Code, Codex, and LangChain) via an async Python client, HTTP API, MCP, or a Core SDK. It supports both basic memory (explicit save/retrieve) and full memory capabilities (automatic extraction and topic-based memory using embedding and generation APIs).

Who it’s for

It is designed for developers and teams who use AI agents for complex, multi-step work that requires hand-offs between different agents or between humans and agents.

Highlights

  • Cross-conversation persistence: Maintains state and progress across different sessions.
  • Broad integration support: Connects with a wide array of agent frameworks including Pydantic AI, LangGraph, and DeepSeek Harness.
  • Handoff management: Specifically organizes current objectives and state to facilitate seamless transitions.
  • Source preservation: Keeps track of sources and earlier revisions of stored information.

Related

  • Project
    akitaonrails/ai-memory

    ai‑memory is a Rust‑based, self‑contained service that gives AI coding assistants a shared, long‑term, markdown‑backed memory. It automatically captures prompts and tool calls via hooks, stores them as git‑versioned markdown pages, and lets any supported agent (Claude Code, Codex, Cursor, Gemini CLI, etc.) retrieve a concise hand‑off on a different machine or after a teammate’s session. By default it works offline with zero LLM calls; optional LLM‑driven “dream” consolidation, vector search, and aging policies are configurable. The project provides native packages, Docker images, and a macOS menu‑bar app, and includes built‑in multi‑user auth, audit logs, and per‑project sharing.

  • Project
    duoduoler-ops/Table-skills

    A collection of AI agent skills designed to maintain project continuity across different chat sessions, primarily through a structured project handoff system.

  • Project
    beilusaiying/always-accompany

    A multi-purpose AI agent framework focused on context and attention mechanisms, providing layered memory and dynamic retrieval for long-term companionship, coding, and work.

  • Project
    JamesShi96/project-butler

    A persistent memory system for AI coding assistants that uses a structured Markdown-based memory stack to maintain project context, rules, and progress across sessions.

  • Project
    CronusL-1141/AI-company

    A shared operating layer for Claude Code and Codex that maintains project memory, tasks, and coordination across AI agent sessions.