Skip to content
Install
One command gets the suite started
Install the main plugin once. It will auto-create the local memory directory, auto-provision bamdra-user-bind, and stage bamdra-memory-vector locally so the full stack is ready immediately.
openclaw plugins install @bamdra/bamdra-openclaw-memoryThe Meaning Of Bamdra
Bamdra is named after the bamboo dragonfly, a simple object lifted by human hands and completed only when it returns.
A bamboo dragonfly rises because a person gives it force. Once airborne, it sees farther and moves beyond the reach of the hand alone. But its value is not separation from people. Its meaning is fulfilled only when it returns with a broader view. That is how we think about AI: launched by human intent, expanded by machine capability, and always brought back to human judgment, human purpose, and human responsibility.
Why teams look at Bamdra now
OpenClawFocused product fit from day one, with a dedicated memory system built specifically for OpenClaw workflows.
SQLite-firstSimple deployment, predictable persistence, and fast retrieval without introducing unnecessary infrastructure.
Workflow changeWhat matters is not just that AI can do something, but that it changes the rhythm, continuity, and leverage of everyday work.
Human returnAI should always come back to human judgment, human goals, and human accountability, even after it has expanded what is possible.
Product Suite
Three plugins, one continuity-first memory stack
Bamdra now ships as a three-plugin public suite. All three plugins can run independently, while the main memory plugin can also auto-provision its identity companion and stage the vector plugin during install.
bamdra-openclaw-memoryThe main runtime for topic continuity, compact context assembly, fact storage, and long-session recovery in OpenClaw.
bamdra-user-bindThe identity and profile plugin. It resolves raw channel identities into stable user boundaries, and it can run on its own or be auto-provisioned by the main memory plugin.
bamdra-memory-vectorThe semantic retrieval and knowledge-base layer. It is auto-staged by the main install so local Markdown knowledge can join recall immediately.
Why the split mattersSeparate repositories, packages, and release roles make deployment clearer, security boundaries safer, and future iteration easier.
Suite Entry
bamdra-openclaw-memory is the main entry point
bamdra-openclaw-memory gives OpenClaw durable memory for long-running sessions. It separates conversation branches, saves reusable facts, refreshes summaries, and keeps the active prompt compact enough to stay practical in production.
`bamdra-openclaw-memory` can run independently. When it is installed through npm, it can also auto-provision bamdra-user-bind if that dependency is not already materialized in the local OpenClaw extensions directory.
bamdra-memory-vector is also staged automatically, so one install already gives you topic continuity, user-aware memory, and a maintainable local knowledge base.
Use Cases
Built for nonlinear work
- Product operations that jump between incidents, planning, and handoff notes.
- Developer copilots that must preserve project constraints across long sessions.
- Knowledge workflows where stable facts, private notes, and shared docs all need to survive restarts and topic drift.
What Changes
Not just deployment, but a new working pattern
- Less repetition when teams return to paused work.
- Stronger continuity across interruptions, detours, and restarts.
- A more durable way to carry decisions, paths, constraints, preferences, and knowledge files through long workflows.
Get Started
Download the current release
Use the one-command install when possible, or start from the packaged release and wire it into OpenClaw with the guidance on this site.