Onboard Existing Context

This workflow helps you onboard existing documentation files (1-3 files) into your context tree. Perfect for READMEs, specs, or any markdown files already in your codebase. For bulk extraction from your entire codebase, see Bootstrap from Codebase.

Prerequisites

ByteRover CLI installed
Project initialized (run brv to start the REPL and initialize your project)
Existing file(s) with documentation or context (markdown, code, config, PDFs, Office documents, etc.)

Step 1: Identify Your Context Source

You likely have existing documentation in your codebase: README.md, ARCHITECTURE.md, SECURITY.md, or similar files containing valuable context. For this example, we’ll use a security.md file:

Security Best Practices:
Never commit .env files - Add to .gitignore
Use strong secrets - Minimum 32 characters for JWT secrets
Rotate secrets regularly - Especially in production
Use secret management - AWS Secrets Manager, HashiCorp Vault
Validate on startup - Fail fast if configuration is wrong

Step 2: Curate to Context Tree

You can curate context using either your coding agent or manually in the REPL.

Via Coding Agent
Commands
Manual REPL

Copy this prompt and paste it into your coding agent’s chat (e.g., Cursor, Claude Code, etc.):

> use brv curate to intelligently organize content from security.md

Your coding agent will read the file and execute:

brv curate "Security Best Practices:
Never commit .env files - Add to .gitignore
Use strong secrets - Minimum 32 characters for JWT secrets
Rotate secrets regularly - Especially in production
Use secret management - AWS Secrets Manager, HashiCorp Vault
Validate on startup - Fail fast if configuration is wrong"

Curate from a file using the -f flag:

brv curate "Security best practices" -f security.md

Run the /curate command directly in the ByteRover REPL with the file reference:

# In the REPL:
/curate "Security best practices from our documentation" @security.md

The @ syntax references files in your codebase (max 5 files per curation).

ByteRover automatically analyzes the content and organizes it into the context tree, detecting relevant domains (likely compliance or code_style) and creating an appropriate topic structure.

Step 3: Review Changes

Verify what was added:

Commands
Manual REPL

brv vc status

# In the REPL:
/vc status

This shows your context tree changes (added, modified, and deleted files). Your context is already stored locally and ready to query.

Step 4: Query to Verify

Test that your context is retrievable using either your coding agent or manually in the REPL.

Via Coding Agent
Commands
Manual REPL

Copy this prompt and paste it into your coding agent’s chat:

> query the context tree for security best practices using brv query

Your coding agent will execute:

brv query "security best practices"

brv query "security best practices"

Run the /query command directly in the ByteRover REPL:

# In the REPL:
/query security best practices

ByteRover will search intelligently, follow relations between topics, and synthesize an answer from your context tree.

What Happened Behind the Scenes

Curate Command: Whether via your coding agent (brv curate) or manually (/curate), ByteRover processed the content from your file.
Intelligent Curation: ByteRover automatically detected relevant domains, organized content hierarchically, and created markdown files with descriptive names in .brv/context-tree/.
Review: You verified the organization with /status. Your context is stored locally and immediately queryable.
Agentic Search: Whether via your coding agent (brv query) or manually (/query), ByteRover reasoned about your question and retrieved relevant context with citations.

Curating Multiple Files

For 2-3 related files, reference them together using either method.

Via Coding Agent
Commands
Manual REPL

Copy this prompt and paste it into your coding agent’s chat:

> read README.md, AUTH.md then use brv curate to add context about the authentication implementation

Your coding agent will read both files and execute brv curate with their combined context.

Reference multiple files using the -f flag:

brv curate "Authentication implementation patterns" -f README.md -f AUTH.md

You can reference up to 5 files per curation.

Use the @ syntax to reference multiple files:

# In the REPL:
/curate "Authentication implementation patterns" @README.md @AUTH.md

You can reference up to 5 files per curation.

ByteRover will organize everything into the appropriate domains and topics.

Cloud Sync (Optional)

Push/pull requires a cloud account. See Local vs Cloud for setup.

To share your context with teammates or back it up to the cloud, stage, commit, and push:

brv vc add .
brv vc commit -m "onboard project context"
brv vc push

/vc add .
/vc commit -m "onboard project context"
/vc push

Next Steps

Bootstrap from Codebase - Bulk context extraction from entire codebase
Query the Context Tree - Effective query patterns and tips
Curate Context with Intelligence - Advanced curation options, interactive vs autonomous modes

Getting Started

LLM Providers

Agent Connectors

Context Tree

Knowledge Sharing

Memory Swarm

Git-Semantic Version Control

Review Changes

BRV Hub

Headless Mode

Daemon-First Architecture

Local Web UI

Reference

FAQs

Security

ByteRover Cookbook

Onboard Existing Context

Prerequisites

Step 1: Identify Your Context Source

Step 2: Curate to Context Tree

Step 3: Review Changes

Step 4: Query to Verify

What Happened Behind the Scenes

Curating Multiple Files

Cloud Sync (Optional)

Next Steps

​Prerequisites

​Step 1: Identify Your Context Source

​Step 2: Curate to Context Tree

​Step 3: Review Changes

​Step 4: Query to Verify

​What Happened Behind the Scenes

​Curating Multiple Files

​Cloud Sync (Optional)

​Next Steps

Prerequisites

Step 1: Identify Your Context Source

Step 2: Curate to Context Tree

Step 3: Review Changes

Step 4: Query to Verify

What Happened Behind the Scenes

Curating Multiple Files

Cloud Sync (Optional)

Next Steps