Releases: MyAgentHubs/aimemo
Releases · MyAgentHubs/aimemo
v0.4.0
aimemo v0.4.0 - OpenClaw Integration Release
Release Date: 2026-02-20
We're excited to announce aimemo v0.4.0, focused on OpenClaw ecosystem integration with per-skill memory isolation and comprehensive Linux support.
🎯 Highlights
OpenClaw Integration
aimemo now provides a complete solution for OpenClaw skills needing persistent memory:
- Per-Skill Memory Isolation: Each skill gets its own isolated database using the
contextparameter - Zero Infrastructure: No Docker, databases, or cloud services required
- Plug-and-Play: Works with OpenClaw's native MCP support
- Complete Documentation: Step-by-step guides, examples, and workflow diagrams
Why This Matters:
OpenClaw's native memory system has known limitations ("remembers everything but understands none"). aimemo solves this with:
- BM25 full-text search + importance scoring
- Per-skill isolation prevents memory pollution
- Lightweight (single binary) perfect for server deployments
Linux First-Class Support
- One-Line Install:
curl -sSL https://raw.githubusercontent.com/MyAgentHubs/aimemo/main/install.sh | bash - Native Binaries: Static binaries for amd64 and arm64
- Zero Dependencies: Pure Go SQLite (modernc.org/sqlite)
- Server-Ready: Perfect for OpenClaw deployments on Linux VPS
AI Agent Documentation
- llms.txt: Machine-readable documentation following llmstxt.org spec
- Optimized for AI agents to understand aimemo capabilities
- Includes OpenClaw integration instructions
🚀 New Features
Documentation
- OpenClaw Integration Guide - Complete setup in 5 minutes
- OpenClaw Workflow - Architecture deep-dive with ASCII diagrams
- Example Skill - Full GitHub PR reviewer with memory
- llms.txt - AI-readable project documentation
Installation
- Linux Install Script (
install.sh) - One-line installation with auto-detection- Supports amd64 and arm64 architectures
- Automatic latest release download
- Friendly post-install instructions
- Error handling and rollback
Examples
- GitHub PR Reviewer Skill - Complete OpenClaw skill demonstrating:
- Code style preference learning
- Pattern recognition across sessions
- User feedback incorporation
- Progressive knowledge building
📝 Improvements
README Updates
- OpenClaw Quick Start section in both English and Chinese
- Installation options: Added Linux one-liner alongside Homebrew
- Per-skill isolation explanation with file system layout
- Comparison table vs other memory solutions (Cognee, memsearch, Supermemory)
Architecture
- Confirmed CGO_ENABLED=0 builds for maximum portability
- Verified context parameter creates isolated databases correctly
- Tested on Linux amd64/arm64 and macOS arm64
🔧 Technical Details
Memory Isolation
Skills use the context parameter for isolation:
// In SKILL.md instructions:
memory_context({context: "my-skill"})
memory_store({context: "my-skill", entities: [...]})
memory_search({context: "my-skill", query: "..."})Result:
workspace/.aimemo/
├── memory.db # Default/shared
├── memory-my-skill.db # Isolated for my-skill
└── memory-other-skill.db # Isolated for other-skill
Platform Support Matrix
| Platform | Architecture | Status |
|---|---|---|
| Linux | amd64 | ✅ Tested |
| Linux | arm64 | ✅ Tested |
| macOS | amd64 (Intel) | ✅ Tested |
| macOS | arm64 (Apple Silicon) | ✅ Tested |
| Windows | amd64 |
📚 Use Cases
For OpenClaw Users
Perfect for skills that need to:
- Learn user preferences over time
- Remember project-specific patterns
- Track progress across sessions
- Build knowledge bases incrementally
Example skills:
- Code reviewers (style preferences, common issues)
- Notification managers (timing, format, dedup)
- Automation tools (templates, priority rules)
- Documentation generators (style guides, TOC patterns)
For AI Coding Client Users
Works seamlessly with:
- Claude Code
- Cursor
- Windsurf
- Cline (VS Code)
- Continue (VS Code / JetBrains)
- Zed
🔄 Migration Guide
From v0.3.0
No breaking changes. Existing installations continue to work.
Optional enhancement for OpenClaw users:
- Update to v0.4.0:
brew upgrade aimemoor re-run install script - Add
contextparameter to SKILL.md instructions - Each skill will automatically get isolated memory
From Other Memory Solutions
From Cognee/memsearch/Supermemory:
- Install aimemo:
curl -sSL https://raw.githubusercontent.com/MyAgentHubs/aimemo/main/install.sh | bash - Initialize:
aimemo init - Update SKILL.md to use aimemo MCP tools
- Optional: Import existing data via
aimemo import
🐛 Bug Fixes
None in this release (focused on features and documentation).
🙏 Acknowledgments
Thanks to the OpenClaw community for feedback on memory system challenges, and to the MCP spec maintainers for the excellent protocol.
📦 Installation
Linux/macOS (Recommended)
curl -sSL https://raw.githubusercontent.com/MyAgentHubs/aimemo/main/install.sh | bashmacOS via Homebrew
brew install MyAgentHubs/tap/aimemoManual Download
Download from GitHub Releases:
aimemo_0.4.0_linux_amd64.tar.gzaimemo_0.4.0_linux_arm64.tar.gzaimemo_0.4.0_darwin_amd64.tar.gzaimemo_0.4.0_darwin_arm64.tar.gz
🔗 Links
- GitHub: https://github.com/MyAgentHubs/aimemo
- Documentation: https://github.com/MyAgentHubs/aimemo#readme
- OpenClaw Integration: https://github.com/MyAgentHubs/aimemo/blob/main/docs/openclaw-integration.md
- Issues: https://github.com/MyAgentHubs/aimemo/issues
- Discussions: https://github.com/MyAgentHubs/aimemo/discussions
🚧 Coming Soon (v0.5.0)
Planned features:
- ClawHub skill listing
- Memory migration tools
- Enhanced search operators
- Multi-project memory views
- Dashboard/GUI (optional)
📄 Full Changelog
See CHANGELOG.md for detailed changes.
Upgrade today and give your OpenClaw skills the memory they deserve!
curl -sSL https://raw.githubusercontent.com/MyAgentHubs/aimemo/main/install.sh | bash