Preserve coding-agent context by sandboxing bulky tool output and retrieving only relevant session state with Context Mode

Use Context Mode when a coding agent keeps burning context on large tool outputs or loses its place after compaction. It wraps tool-heavy workflows with sandboxed execution, indexed session history, and targeted retrieval so the agent can keep working without reloading raw data into the prompt.

Prerequisites

Supported coding agent runtime with MCP or plugin support; Node.js; local SQLite storage

Installation

Use the upstream install or setup path that matches your environment:

• npm install -g context-mode
• git clone https://github.com/mksglu/context-mode.git
• npm run install:openclaw
• npm run install:openclaw -- /path/to/openclaw-state

Requirements and caveats from upstream:

• Prerequisites: Claude Code v1.0.33+ (claude --version). If /plugin is not recognized, update first: brew upgrade claude-code or npm update -g @anthropic-ai/claude-code.
• Prerequisites: Node.js >= 22.5 (or Bun), Gemini CLI installed.
• Prerequisites: Node.js >= 22.5 (or Bun), VS Code with Copilot Chat v0.32+.

Basic usage or getting-started notes:

• Platforms are grouped by install complexity. Hook-capable platforms get automatic routing enforcement. Non-hook platforms need a one-time routing file copy.

• <details open>
• <summary><strong>Claude Code</strong> — plugin marketplace, fully automatic</summary>

• Source: https://github.com/mksglu/context-mode

• Extracted from upstream docs: https://raw.githubusercontent.com/mksglu/context-mode/HEAD/README.md

Documentation

• https://context-mode.com

Source

• Agent Skill Exchange