Store and link agent memories as structured records in RushDB.
npx skills add https://github.com/rush-db/rushdb --skill rushdb-agent-memory
Please help me install this Agent Skill. Skill: rushdb-agent-memory Source: https://github.com/rush-db/rushdb/tree/main/packages/skills/rushdb-agent-memory Command: npx skills add https://github.com/rush-db/rushdb --skill rushdb-agent-memory
RushDB Agent Memory provides a persistent, structured memory layer for AI agents, enabling durable storage of session data, decisions, and context. It unifies memory across conversations by constructing a connected graph of memories and relationships, replacing disparate vector stores and key-value caches with a single ACID-safe store.
Store a memory with a labeled data object and recall it later by semantic meaning.
High-intent search queries and answers about installing and using this skill.
Persistent memory is achieved by storing session data and decisions as structured records in a graph database, enabling semantic recall and relationship traversal across multiple sessions. This approach replaces fragmented key-value caches with a unified ACID-safe store.
Semantic recall is handled by storing labeled data objects as typed records and querying them by meaning. This allows agents to retrieve relevant context without relying on exact keyword matches.
Yes, an MCP server and specific client tooling are required to create records, bulk create data, query by meaning, and traverse relationships when using this graph-backed agent memory approach.
Graph traversal improves context by auto-linking nested data into a connected graph of memories and relationships, replacing disparate vector stores with a single ACID-safe store that supports complex relationship queries.
Yes, memories can be stored as typed records such as SESSION, DECISION, or ENTITY within a single unified memory layer. This allows diverse data types to be auto-linked and queried together.
matthewdeavesRegister a new Classic Mac in the MCP server and verify FTP connectivity.
nmdraIndex and query Obsidian markdown vaults with ChromaDB for semantic retrieval and graph analysis.
webrenewManage AI agent memories via CLI with SQLite and cloud sync.
brenordvModel NoSQL database systems using query-driven structures and horizontal scalability.
