AVIDS2/memorix — Open-source cross-agent memory layer for coding agents via MCP.

Description and use

**Brief Summary (2‑3 sentences)**
AVIDS2 / memorix is an open‑source, TypeScript‑based memory layer that lets coding agents share state through the MCP (Memory‑Coordination‑Protocol). It works out‑of‑the‑box with a wide range of agents—including Cursor, Claude Code, Codex, Gemini CLI, GitHub Copilot, and others—turning isolated prompts into repeatable, orchestrated workflows.

**Value**

• **Unified Agent Memory:** Provides a common “brain” that persists context across multiple agents, eliminating the need to rebuild state handling for each tool.
• **Workflow Repeatability:** By standardising how agents read/write memory, teams can encode complex multi‑agent pipelines (e.g., code generation → static analysis → refactoring) that can be version‑controlled and reused.
• **Broad Compatibility:** Native adapters for the most popular coding assistants mean you can plug memorix into existing stacks without rewriting code.

**Practical Adoption Path**

1. **Prototype:** Clone the repo, run the provided CLI or import the SDK into a sandbox project, and connect a single agent (e.g., Cursor) to verify basic read/write operations.
2. **Integrate Additional Agents:** Add the other supported adapters (Claude Code, Gemini CLI, etc.) using the documented API endpoints; no language changes are required because the layer is language‑agnostic and communicates over HTTP/JSON.
3. **Define Memory Schemas:** Create JSON‑schema definitions for the data you want to share (e.g., AST snapshots, lint reports) and register them with memorix’s metadata service.
4. **Orchestrate Pipelines:** Use a simple orchestration script (or an existing workflow engine) to sequence agents, letting each step read the prior step’s memory and write its own results.
5. **Scale & Monitor:** Deploy memorix as a containerized microservice behind your CI/CD pipeline; leverage its built‑in metrics and logging to monitor latency and storage usage.

**Production Readiness**

• **Activity & Adoption:** 514 ★, 39 forks, recent commits (as of 2026‑06‑23), and active usage across several major coding agents indicate a healthy community.
• **Technical Maturity:** Written in TypeScript with a clear API/SDK/CLI surface, extensive topic tagging, and straightforward integration points.
• **Risk Profile:** No glaring licensing or security red flags have been identified, though a final audit of the license terms and a vulnerability scan are recommended before enterprise rollout.

Overall, memorix is a high‑readiness OSS component suitable for pilots and, with standard security vetting, for production‑grade multi‑agent coding pipelines.

Problem

AVIDS2/memorix helps turn isolated prompts and tools into repeatable agent workflows.

Outcome

High for an OSS candidate: recent activity, adoption, and ecosystem signals are strong enough for a serious pilot.

Workflow

• coordinate multi-agent workflows
• add tool-use pipelines
• standardize agent memory

Evidence

Open-source cross-agent memory layer for coding agents via MCP. Compatible with Cursor, Claude Code, Codex, Windsurf, Gemini CLI, GitHub Copilot, Kiro, OpenCode, Antigravity, and Trae.

Source