MCP App StoreMCP App Store
from brainx-the-first-brain-for-openclaw31
Persistent vector-memory engine for OpenClaw agents providing semantic search, durable recall, and a plugin bridge for JIT recall and working memory.
BrainX V6 provides durable, semantic agent memory built on PostgreSQL + pgvector and OpenAI embeddings. It lets agents store, search, curate, and inject contextual memories across sessions, with lifecycle controls (hot/warm/cold), deduplication, verification states, and tools for backup, restore, diagnostics, and evaluations. A companion OpenClaw plugin exposes conservative runtime features—wiki digest, JIT recall, working memory, and pre-action advisories—while encouraging conservative default settings for privacy and safety.
Use BrainX when an agent must remember cross-session facts, maintain a knowledge vault, or perform semantic retrieval to enrich reasoning. It is appropriate for agents that need persistent context (project history, recurring SOPs, user preferences) and for setups that require auditable memory states and PII controls. Start conservative (JIT recall off) until retention and privacy rules are configured.
Best suited for agents with shell/CLI and DB access (OpenClaw agents, Claude Code/Copilot-style automation, CLI-driven Copilot/Codex workflows).
BrainX V6 is a persistent vector-memory engine for OpenClaw agents using PostgreSQL+pgvector+OpenAI embeddings. It offers 41 scripts covering memory distillation, dedup, contradiction detection, backup/restore, cross-agent learning, and more. Shell scripts ran; backup and cleanup worked correctly while restore and weekly-backup needed missing paths. All 37 JS scripts were skipped by the runner, requiring a Node+DB environment to test properly. SKILL.md is thorough with good trust model and conservative defaults, but the scripts directory is a monolithic flat dump with no clear entry-point separation.
node/npm (for .js scripts)psql/pg_dump (for DB ops)pgvector extensionSecurity: No hardcoded credentials, no curl|bash, no exfiltration. Backup script sources .env via grep+xargs which is a minor shell injection risk if .env contains special chars (-12). Scripts are read-only within DB and local filesystem. Code quality: Scripts are functional but inconsistent — some have good arg parsing, others are minimal. Error handling varies. Architecture: Flat scripts/ directory with 41 files, no sub-grouping. SKILL.md is well-structured but scripts lack clear contract documentation. No references/ directory. Usefulness: Solves a real problem (durable agent memory) with comprehensive tooling. Broad audience among OpenClaw users but requires significant setup (Postgres, pgvector, OpenAI).
Sync Production Database to Development
Run a safe, scripted workflow to download production database dumps and restore them into a development environment (download-only and restore options).
Caveman Compress
Compress natural-language memory files into a compact 'caveman' format to save LLM tokens while preserving code, links, and structure.
Sandbox0 Integration
Guidance and templates for integrating Sandbox0 sandboxing into AI agents — CLI/SDK patterns, templates, volumes, network policy, and deployment choices.
Unitree Robot Controller
Control and command Unitree robots (GO1/GO2/G1/H1) via OpenClaw: initialization, basic motion commands, and sensor integrations.
Go Data Structures
Authoritative guidance on choosing and using Go built-in and standard-library data structures, with practical best practices for slices, maps, arrays, container
Memory Search
Search past coding sessions and observations using natural-language queries to retrieve decisions, notes, and context.
Easy OpenClaw Config Guide
Interactive configuration and optimization guide for OpenClaw—walks users through testing, three rounds of channel-specific tuning, and optional channel onboard
Cashu Wallet
Privacy-focused Bitcoin wallet skill enabling sending/receiving via Lightning and Cashu ecash tokens, with multi-mint support and basic wallet management comman