Store and retrieve cross-project memories via local semantic search.
npx skills add https://github.com/ybonda/memo --skill memo-ybonda
Please help me install this Agent Skill. Skill: memo Source: https://github.com/ybonda/memo/tree/main Command: npx skills add https://github.com/ybonda/memo --skill memo-ybonda
AI coding agents lose all memory between sessions, forcing you to repeatedly explain architecture decisions, bug patterns, and project context. This Skill gives your agent a persistent, semantic memory layer that lives on your machine, so it remembers everything across conversations and projects without you having to restate it.
Use the memo skill to save an important insight by telling your agent to remember it, and later ask your agent what it knows about any topic to instantly retrieve relevant memories.
High-intent search queries and answers about installing and using this skill.
Persistent memory for AI coding agents is provided by a local semantic search layer that stores context in SQLite with vector embeddings, ensuring architecture decisions and bug patterns are recalled across sessions without manual repetition.
Semantic search for local project context works by generating vector embeddings for stored memories, enabling retrieval of architecture decisions and bug patterns by meaning rather than exact keyword matching.
Yes, Obsidian can browse AI agent memories because every memory is rendered as a Markdown file in a local vault, providing a searchable interface for reviewing project context and architecture decisions.
Yes, cross-project memory recall requires a local MCP server and SQLite with vector extensions to manage semantic search and store architecture decisions, bug patterns, and project context.
Duplicate memories are prevented by detecting exact duplicates via SHA256 hashing and identifying near-duplicate project context entries through cosine similarity checks on their vector embeddings.
The best way to share bug pattern insights across projects is a cross-project knowledge base with local semantic search, making insights from one project instantly available in all others through a persistent memory layer.
AlexiosBluffMaraIndex and search local markdown notes with hybrid retrieval.
ncz-osStore and retrieve persistent memories across agent sessions via API.
lvndryAutomate Obsidian note creation, retrieval, and vault organization via obsidian-cli.
dglijin-ossStore and search memories locally using LanceDB and Ollama embeddings.
