Releases·Tool Release
shared-memory-vault: Markdown-based memory governance engine lets multiple AI agents share one cloud knowledge base
A GitHub project called shared-memory-vault ships a memory governance engine for multi-agent cloud knowledge bases. It stores everything in pure Markdown, adds nine governance automations, and has zero dependencies, aiming to let every agent and device share one consistent knowledge base.
27 engagement1 sourceSat, Sep 5, 2026, 05:15 AM
AI Overview
- shared-memory-vault turns a cloud-synced Markdown directory into a shared agent knowledge base with nine governance automations, zero dependencies, and no database
- Agents coordinate via a session-blackboard convention (CONTEXT.md), letting any agent resume work from where another left off
- Consistency uses a single source of truth with conflict handling via timestamps plus .bak files and incremental updates
- The knowledge base works across Mac, Windows, Linux, and mobile with no server or vendor lock-in
Entities
Related
Social·Community
What are you using for a shared, agent-native knowledge base across your team?
A developer surveys shared knowledge-base tools that both humans and AI agents can read and write, covering ByteRover, Basic Memory, and the open-source OpenLore. OpenLore serves existing Markdown files over SSH or MCP with role-based access and no vector database, illustrating a file-first alternative to indexed memory pipelines.
A developer surveys shared knowledge-base tools that both humans and AI agents can read and write, covering ByteRover, Basic Memory, and the open-source OpenLore. OpenLore serves existing Markdown files over SSH or MCP with role-based access and no vector database, illustrating a file-first alternative to indexed memory pipelines.
4 engagement1 sourceredditTue, Oct 6, 2026, 03:30 PM
Products·launched
Shared memory layer for AI and team context
ContextVault is a shared memory layer that stores prompts, coding conventions, architectural decisions, and other context for AI tools. It prevents fragmentation of information across ChatGPT, Claude, markdown files, and internal docs, enabling teams to reuse context and avoid solving the same problems independently.
ContextVault is a shared memory layer that stores prompts, coding conventions, architectural decisions, and other context for AI tools. It prevents fragmentation of information across ChatGPT, Claude, markdown files, and internal docs, enabling teams to reuse context and avoid solving the same problems independently.
2 engagement1 sourcehackernewsMon, Jul 13, 2026, 11:22 PM
Products·launched
MemBridge: shared memory layer for AI coding tools
MemBridge is a local daemon that automatically captures decisions made in AI coding sessions (Claude Code, Codex) and writes them into CLAUDE.md and AGENTS.md files for persistent team memory. It solves the problem of context window forgetting across team members, ensuring agents don't waste tokens on already-resolved issues.
MemBridge is a local daemon that automatically captures decisions made in AI coding sessions (Claude Code, Codex) and writes them into CLAUDE.md and AGENTS.md files for persistent team memory. It solves the problem of context window forgetting across team members, ensuring agents don't waste tokens on already-resolved issues.
2 engagement1 sourceredditMon, Jul 27, 2026, 04:54 PM
Products·launched
Memoars: Encrypted shared memory layer for AI assistants via MCP
Memoars is an encrypted memory layer that lets multiple AI assistants share a single, user-owned memory store. It uses MCP (Model Context Protocol) for read/write access, encrypts content locally with XChaCha20-Poly1305 and Argon2id key derivation, and aims to solve context and knowledge fragmentation across AI agents.
Memoars is an encrypted memory layer that lets multiple AI assistants share a single, user-owned memory store. It uses MCP (Model Context Protocol) for read/write access, encrypts content locally with XChaCha20-Poly1305 and Argon2id key derivation, and aims to solve context and knowledge fragmentation across AI agents.
7 engagement2 sourceshackernews, redditThu, Aug 6, 2026, 12:29 PM
Products·launched
Callosium: shared memory folder for AI assistants
Callosium is a local folder of plain markdown files that multiple AI assistants (Claude, ChatGPT, Cursor, and 23 others) can read and write to. It solves the problem of each AI having separate, siloed memory by giving them a single shared brain that the user owns. Users teach one AI something and all connected AIs know it, with answers citing the exact note used.
Callosium is a local folder of plain markdown files that multiple AI assistants (Claude, ChatGPT, Cursor, and 23 others) can read and write to. It solves the problem of each AI having separate, siloed memory by giving them a single shared brain that the user owns. Users teach one AI something and all connected AIs know it, with answers citing the exact note used.
1 engagement1 sourceredditFri, Jul 17, 2026, 03:02 PM