Your context.
Every agent.

A thin wrapper around Amazon Bedrock Knowledge Bases. Self-host it now.

Connects the tools your team already uses
GitHub
Notion
Google Docs
Google Sheets
Google Slides

Self-host now.

Bedrock Knowledge Bases and per-brain MCP. You run the CLI in your AWS.

  1. 01

    Install the CLI

    Install the published CLI, then you're ready to init.

    npm i -g context101-cli
  2. 02

    Init a space

    It prompts for a space name.

    context101 init

Sources in. Agents retrieve.

Connect your sources. Every agent retrieves the same docs.

  1. 01

    Connect your sources

    Connect Google Drive, Notion, GitHub, or uploads. Agents retrieve the same docs.

  2. 02

    Retrieve through MCP

    Cursor, Claude, Devin, or your own agents hit per-brain MCP. Same KB. Raw retrieve.

Isolated brains on AWS.

Each brain is its own bucket, Bedrock KB, vector index, and MCP token. The admin UI manages brains, sources, and the review queue.

  1. Isolated brains

    Own S3 bucket, own Bedrock KB, own vector index, own MCP bearer token.

  2. Admin UI

    Create brains, connect sources, and manage access.

  3. Team access

    Signed-in teammates can manage brains, sources, and the review queue.

Raw retrieve. Human review.

Agents search the source docs through Bedrock. Suggestions wait for a person.

Search it

Bedrock retrieve over the raw docs. MCP serves those chunks to any agent.

Review it

Agents can suggest a change. A human approves it in the queue before it lands in the KB.