Releases: contextpilot-dev/memorypilot
Releases · contextpilot-dev/memorypilot
v0.1.1 - Complete daemon management and MCP fixes
What's New
Features
- Daemon management: Start/stop/status with PID file tracking
- Background mode:
memorypilot daemon start -bruns detached - MCP remember fixed: Now actually saves memories to database
- Embedding generation: Memories get vector embeddings for semantic search
- Hybrid search: MCP recall uses semantic + keyword search
- Cross-platform: Works on macOS, Linux, and Windows
Usage
# Initialize
memorypilot init
# Start daemon in background
memorypilot daemon start -b
# Check status
memorypilot daemon status
# Remember something
memorypilot remember "Always use transactions for DB ops" --type decision
# Search memories (semantic + keyword)
memorypilot recall "database best practices"
# Stop daemon
memorypilot daemon stopInstallation
# macOS (Apple Silicon)
curl -L https://github.com/contextpilot-dev/memorypilot/releases/download/v0.1.1/memorypilot-darwin-arm64 -o memorypilot
chmod +x memorypilot && sudo mv memorypilot /usr/local/bin/
# Linux
curl -L https://github.com/contextpilot-dev/memorypilot/releases/download/v0.1.1/memorypilot-linux-amd64 -o memorypilot
chmod +x memorypilot && sudo mv memorypilot /usr/local/bin/Requirements
- Ollama for semantic search:
ollama pull nomic-embed-text - Optional:
ollama pull llama3.2for automatic memory extraction
v0.1.0 - Initial Release
MemoryPilot v0.1.0
One memory. Every AI. Zero repetition.
Part of the ContextPilot family.
Features
- 🧠 Passive capture (git, file, terminal watchers)
- 🔍 Semantic search with vector embeddings
- 🤖 AI-powered memory extraction (via Ollama)
- 🔌 MCP server for Claude Code / OpenClaw / Windsurf
- 📦 Single binary, no dependencies
Installation
curl -fsSL https://contextpilot.dev/memorypilot/install.sh | shOr with Go:
go install github.com/contextpilot-dev/memorypilot@latestQuick Start
memorypilot init
memorypilot daemon start
memorypilot recall "auth patterns"Requirements
- For AI features: Ollama with llama3.2 and nomic-embed-text models