Skip to content

BlackIron-Technologies/Neurotrace_c

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

20 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

NeuroTrace MCP

Persistent Memory for AI Agents That Lives With Your Code

Persistent project memory for developers and coding agents inside the IDE.


See NeuroTrace In Action

Your agent starts every new chat with amnesia.

NeuroTrace gives developers and coding agents a local-first, code-linked memory layer that survives across chats and sessions.

NeuroTrace helps agents and developers:

  • Save high-signal memory during real work
  • Retrieve project context before coding
  • Reuse decisions, tasks, and insights across sessions
  • Keep memory attached to the codebase instead of the chat

Built for long-running codebases, debugging, planning, architecture work, and agent-assisted development.


Quick Start

  1. Install NeuroTrace from the VS Code Marketplace.
  2. Open the NeuroTrace sidebar.
  3. Download the platform backend if prompted.
  4. Click Initialize Database.
  5. Let NeuroTrace generate the MCP setup and host-specific guidance for your workspace.
  6. Start a fresh agent chat and use NeuroTrace in that repo.

Generated setup includes:

  • .neurotrace/mcp/README.md with ready-to-use MCP templates
  • .cursor/rules/neurotrace.mdc for Cursor when supported
  • .github/copilot-instructions.md for Copilot workflows in VS Code
  • automatic Codex MCP rebinding to the active workspace

More details: walkthrough/init.md


How It Works

NeuroTrace gives coding agents a persistent memory layer they can query before acting and update after the work is done.

Typical workflow:

  1. The agent receives a task.
  2. It checks NeuroTrace for prior decisions, tasks, and related context.
  3. It works with project history available inside the workspace.
  4. It saves the outcome as a structured memory.
  5. It links important context for future sessions.

Key MCP capabilities include:

  • Create and edit structured memories
  • Search by keyword or semantic meaning
  • Retrieve file-scoped context
  • Discover related memories and graph connections
  • Check database readiness before work begins

What NeuroTrace Sets Up

After initialization, NeuroTrace prepares the workspace for both humans and agents:

  • Creates the local .neurotrace workspace directory
  • Stores the project memory database locally
  • Auto-configures supported agent hosts where possible
  • Generates MCP setup files under .neurotrace/mcp/
  • Generates host-specific guidance such as .cursor/rules/neurotrace.mdc or .github/copilot-instructions.md

For client-specific setup and troubleshooting:


Interactive Graph

NeuroTrace also includes a graph view for exploring connected project memory.

Graph


Privacy And Security

For transparency, security-critical code is open source in our GitHub repository.

  • 100% local-first storage
  • Encrypted database support (SQLCipher3, AES-256 at rest)
  • Optional anonymous telemetry
  • Works offline

See also:


Support

License

NeuroTrace uses a hybrid licensing model: security-critical code is open source, while premium features remain proprietary. See LICENSE.md for complete terms.


Built to remember.
© 2026 BlackIron Technologies Ltd. All rights reserved.

About

NeuroTrace is a local-first memory system for AI coding agents and developers inside your IDE.

Resources

License

Stars

Watchers

Forks

Packages

 
 
 

Contributors