Memory Lancedb Pro
Maintain and extend the OpenClaw memory-lancedb-pro plugin when you change retrieval, scopes, embeddings, CLI tools, or debug noisy long-term agent memory.
Overview
memory-lancedb-pro is an agent skill most often used in Build (also Operate debugging) that guides development and troubleshooting of the OpenClaw LanceDB hybrid memory plugin.
Install
npx skills add https://github.com/win4r/memory-lancedb-pro-skill --skill memory-lancedb-proWhat is this skill?
- Hybrid retrieval pipeline: vector search, BM25, RRF fusion, cross-encoder reranking, and multi-stage scoring
- Multi-scope memory isolation plus noise filtering and adaptive retrieval for long-term OpenClaw sessions
- Management CLI and agent tools for operating LanceDB-backed memory beyond stock memory-lancedb
- JSONL session distillation pipeline and migration guidance between memory backends
- TypeScript ESM plugin entry at index.ts running on Node via OpenClaw Gateway
Adoption & trust: 620 installs on skills.sh; 47 GitHub stars; 3/3 security scanners passed (skills.sh audits).
What problem does it solve?
Your OpenClaw agent memory is noisy, poorly scoped, or hard to change because the stock LanceDB plugin lacks the hybrid retrieval and CLI controls you need.
Who is it for?
OpenClaw operators and plugin contributors who own TypeScript memory code and need hybrid Vector+BM25 behavior with reranking and scope isolation.
Skip if: Builders who only want copy-paste RAG in a generic app without OpenClaw, or teams unwilling to run Node gateway plugins.
When should I use this skill?
Developing, debugging, or upgrading memory-lancedb-pro; changing retrieval, embeddings, scopes, tools, CLI, distillation, or migrations; troubleshooting memory quality.
What do I get? / Deliverables
You can modify retrieval stages, scopes, embeddings, and CLI safely and diagnose recall or duplication issues with a documented architecture map.
- Pipeline or scope code changes
- CLI or agent tool updates
- Migration or distillation fixes with validated recall
Recommended Skills
Journey fit
Spans multiple journey phases - primary shelf plus alternate fits below.
The skill documents plugin architecture and pipeline code paths—the primary shelf is Build agent-tooling for OpenClaw memory systems. Hybrid Vector+BM25 retrieval, reranking, scopes, and agent tools are core agent-runtime tooling rather than generic app frontend work.
Where it fits
Add a new embedding provider flag without breaking existing scoped memory tables.
Expose a new agent tool that queries reranked memories from the management CLI.
Trace why BM25 hits overpower vector matches after a tokenizer change.
Tune adaptive retrieval thresholds after JSONL distillation quality regresses.
How it compares
OpenClaw plugin maintenance playbook—not a hosted memory SaaS or a one-shot ‘enable long-term memory’ prompt.
Common Questions / FAQ
Who is memory-lancedb-pro for?
Developers maintaining the memory-lancedb-pro OpenClaw plugin who touch retrieval pipelines, embeddings, scopes, tools, or session distillation.
When should I use memory-lancedb-pro?
In Build when adding features to agent memory; in Operate when fixing low recall, noise, duplicates, or migrating between memory backends.
Is memory-lancedb-pro safe to install?
It guides code that runs inside your gateway and stores embeddings; review Security Audits on this page and audit scope/access before production agents.
SKILL.md
READMESKILL.md - Memory Lancedb Pro
# memory-lancedb-pro Plugin Maintenance Guide ## Overview **memory-lancedb-pro** is an enhanced long-term memory plugin for [OpenClaw](https://github.com/openclaw/openclaw). It replaces the built-in `memory-lancedb` plugin with advanced retrieval capabilities, multi-scope memory isolation, and a management CLI. **Repository**: https://github.com/win4r/memory-lancedb-pro **License**: MIT | **Language**: TypeScript (ESM) | **Runtime**: Node.js via OpenClaw Gateway ## Architecture ``` ┌─────────────────────────────────────────────────────────┐ │ index.ts (Entry Point) │ │ Plugin Registration · Config Parsing · Lifecycle Hooks │ └────────┬──────────┬──────────┬──────────┬───────────────┘ │ │ │ │ ┌────▼───┐ ┌────▼───┐ ┌───▼────┐ ┌──▼──────────┐ │ store │ │embedder│ │retriever│ │ scopes │ │ .ts │ │ .ts │ │ .ts │ │ .ts │ └────────┘ └────────┘ └────────┘ └─────────────┘ │ │ ┌────▼───┐ ┌─────▼──────────┐ │migrate │ │noise-filter.ts │ │ .ts │ │adaptive- │ └────────┘ │retrieval.ts │ └────────────────┘ ┌─────────────┐ ┌──────────┐ │ tools.ts │ │ cli.ts │ │ (Agent API) │ │ (CLI) │ └─────────────┘ └──────────┘ ``` ## File Reference (Quick Navigation) | File | Purpose | Key Exports | |------|---------|-------------| | `index.ts` | Plugin entry point. Registers with OpenClaw Plugin API, parses config, mounts lifecycle hooks | `memoryLanceDBProPlugin` (default), `shouldCapture`, `detectCategory` | | `openclaw.plugin.json` | Plugin metadata + full JSON Schema config with `uiHints` | — | | `package.json` | NPM package. Deps: `@lancedb/lancedb`, `openai`, `@sinclair/typebox` | — | | `cli.ts` | CLI: `memory-pro list/search/stats/delete/delete-bulk/export/import/reembed/migrate` | `createMemoryCLI`, `registerMemoryCLI` | | `src/store.ts` | LanceDB storage layer. Table creation, FTS indexing, CRUD, vector/BM25 search | `MemoryStore`, `MemoryEntry`, `loadLanceDB` | | `src/embedder.ts` | Embedding abstraction. OpenAI-compatible API, task-aware, LRU cache | `Embedder`, `createEmbedder`, `getVectorDimensions` | | `src/retriever.ts` | Hybrid retrieval engine. Full scoring pipeline | `MemoryRetriever`, `createRetriever`, `DEFAULT_RETRIEVAL_CONFIG` | | `src/scopes.ts` | Multi-scope access control | `MemoryScopeManager`, `createScopeManager` | | `src/tools.ts` | Agent tool definitions: `memory_recall/store/forget/update/stats/list` | `registerAllMemoryTools` | | `src/noise-filter.ts` | Noise filter for low-quality content | `isNoise`, `filterNoise` | | `src/adaptive-retrieval.ts` | Skip retrieval for greetings, commands, emoji | `shouldSkipRetrieval` | | `src/migrate.ts` | Migration from legacy `memory-lancedb` | `MemoryMigrator`, `createMigrator` | | `scripts/jsonl_distill.py` | JSONL session distillation script (Python) | — | ## Core Subsystem Reference For detailed deep-dives into each subsystem, r