Step-by-step guides for upgrading between NeuralMind versions.
| From | To | Effort | Breaking Changes |
|---|---|---|---|
| v0.3.x | v0.4.x | 5 min | Minor (CLI args) |
| v0.4.x | v0.4.y | 1 min | None |
| v0.4.x | v0.5.0 | 30 min | Yes (MCP server) |
| v0.5.x | v1.0.0 | 1 hour | Possibly (API) |
| Effort: 5 minutes | Risk: Very Low |
These are bug fixes and security patches—no breaking changes.
# Upgrade pip package
pip install --upgrade neuralmind
# Verify version
neuralmind --version
# Rebuild index (recommended after patch)
neuralmind build . --force
After upgrade:
| Effort: 10 minutes | Risk: Low | Breaking: Minor |
CLI:
graphify build → graphify update (graphify 1.2+ only)--old-format flag deprecated (use --json instead)--backend option for embedding selectionIndexes:
1. Backup current index
cd /path/to/project
cp -r neuralmind.db neuralmind.db.backup
cp -r .neuralmind .neuralmind.backup
2. Upgrade NeuralMind
pip install --upgrade neuralmind
# Verify version
neuralmind --version
# Output: neuralmind v0.4.x
3. Upgrade graphify (if needed)
cd graphify
git pull origin main
pip install -e .
# Verify version
graphify --version
# Should be v1.2.0+
4. Update code graph
cd /path/to/project
graphify update .
# Verify graph updated
ls -la graphify-out/graph.json
5. Rebuild NeuralMind index
neuralmind build . --force
# This regenerates embeddings with v0.4 format
# Takes ~same time as original build
6. Verify everything works
# Test basic commands
neuralmind stats .
neuralmind wakeup .
neuralmind query . "What is the main purpose of this project?"
# Compare results with backup if needed
7. Clean up backup (if confident)
rm -rf neuralmind.db.backup .neuralmind.backup
# Uninstall v0.4
pip uninstall neuralmind
# Reinstall v0.3
pip install neuralmind==0.3.x
# Restore backup
rm -rf neuralmind.db .neuralmind
cp -r neuralmind.db.backup neuralmind.db
cp -r .neuralmind.backup .neuralmind
# You're back to v0.3
| Effort: 30 minutes | Risk: Medium | Breaking: Yes |
MCP Server:
Deprecations:
New Features:
1. Read Breaking Changes
See: docs/v0.4-to-v0.5-MIGRATION.md
2. Backup Everything
cd /path/to/project
# Backup index
cp -r neuralmind.db neuralmind.db.v0.4.backup
cp -r .neuralmind .neuralmind.v0.4.backup
# Backup configuration
cp neuralmind.toml neuralmind.toml.v0.4.backup
3. Check Python Version
python --version
# Must be 3.10 or higher
# If not, upgrade:
# See: https://python.org/downloads/
4. Upgrade NeuralMind
pip install neuralmind==0.5.0
neuralmind --version
# Output: neuralmind v0.5.0
5. Update Configuration (if using custom config)
# neuralmind.toml - v0.5.0 changes
[build]
# New option: backend specification
backend = "chromadb" # or "postgres", "lancedb"
[backends.postgres]
# New! PostgreSQL support
connection_string = "postgresql://user:pass@host/db"
pool_size = 10
# Deprecated (remove if present):
# embedding_model = "sentence-transformers" # Now specified elsewhere
6. Rebuild Index
neuralmind build . --force
# This rebuilds with v0.5 format
# May take longer due to new optimizations
7. Update MCP Configuration (if using MCP)
Old (v0.4):
{
"mcpServers": {
"neuralmind": {
"command": "neuralmind-mcp",
"args": ["."]
}
}
}
New (v0.5):
{
"mcpServers": {
"neuralmind": {
"command": "neuralmind-mcp",
"args": ["."],
"env": {
"NEURALMIND_VERSION": "0.5.0",
"NEURALMIND_BACKEND": "chromadb"
}
}
}
}
8. Test All Commands
# Test basic commands
neuralmind stats .
neuralmind wakeup .
neuralmind query . "What does this project do?"
# Test MCP (if applicable)
# Start Claude Code and verify tools load
# Test new features
neuralmind backend-list
neuralmind backend-check
9. Verify Performance
neuralmind benchmark .
# Compare against v0.4 baseline
# Should see similar or better token reduction
10. Update Documentation (for your team)
# Create migration guide for your usage
echo "# Upgrading to NeuralMind v0.5" > docs/NEURALMIND_UPGRADE.md
| Issue | Workaround | Status |
|---|---|---|
| Slow first query | First query optimizes cache; subsequent queries fast | Fixed in v0.5.1 |
| pgvector connection timeout | Set NEURALMIND_DB_TIMEOUT=30 |
Use v0.5.2+ |
| Old .neuralmind format causes error | Delete .neuralmind/cache and rebuild |
Normal behavior |
# Uninstall v0.5
pip uninstall neuralmind
# Reinstall v0.4
pip install neuralmind==0.4.x
# Restore backup
rm -rf neuralmind.db .neuralmind neuralmind.toml
cp -r neuralmind.db.v0.4.backup neuralmind.db
cp -r .neuralmind.v0.4.backup .neuralmind
cp neuralmind.toml.v0.4.backup neuralmind.toml
# Rebuild with v0.4
neuralmind build . --force
| Effort: 1 hour | Risk: Low-Medium | Breaking: Minimal |
v1.0.0 is the first stable release with API guarantees.
# Simple upgrade path from v0.5.x
pip install neuralmind==1.0.0
# Rebuild recommended (one-time)
neuralmind build . --force
# No configuration changes needed
Upgrading 10+ projects at once:
#!/bin/bash
# upgrade-all-projects.sh
PROJECTS_ROOT="$HOME/projects"
NEURALMIND_VERSION="0.4.2"
for project_dir in "$PROJECTS_ROOT"/*; do
if [ ! -d "$project_dir" ]; then
continue
fi
echo "Upgrading: $project_dir"
cd "$project_dir"
# Check if neuralmind initialized
if [ ! -d "neuralmind.db" ]; then
echo " → Skipping (not initialized)"
continue
fi
# Backup
cp -r neuralmind.db "neuralmind.db.backup-$(date +%Y%m%d)"
# Rebuild
neuralmind build . --force
# Verify
if neuralmind stats . > /dev/null 2>&1; then
echo " → ✅ Success"
else
echo " → ❌ Failed (rolled back)"
rm -rf neuralmind.db
cp -r "neuralmind.db.backup-$(date +%Y%m%d)" neuralmind.db
fi
done
echo "Upgrade complete!"
Run it:
chmod +x upgrade-all-projects.sh
./upgrade-all-projects.sh
| Version | Status | Python | End of Support |
|---|---|---|---|
| v1.0.x | LTS | 3.10-3.13+ | 2029-01 |
| v0.5.x | Current | 3.10-3.13 | 2026-10 |
| v0.4.x | LTS | 3.10-3.13 | 2026-10 |
| v0.3.x | Maintained | 3.10-3.12 | 2026-04 |
| v0.2.x | EOL | 3.10-3.11 | ❌ 2025-10 |