Products & Announcements

Recall: Persistent Redis Memory for Claude (v1.5)

Model Context ProtocolLLM Context ManagementDeveloper ToolingAI Coding Agents
Added Oct 8, 2025
Link93
Article: PositiveCommunity: NeutralDivisive

Recall adds durable, searchable memory to Claude using Redis, so important context persists across sessions and survives context limits. It offers a comprehensive toolkit—core memory ops, smart context, relationships/graphs, version history/rollback, templates, categories, and advanced search—with workspace isolation and optional global sharing. Strong security guidance, easy setup via npx, and low cost make it suitable for both individual developers and teams.

Key Points

  • Persistent memory for Claude stored in Redis, isolated by workspace with optional global sharing (isolated/global/hybrid modes).
  • Rich toolset: store/search/update/delete, smart context retrieval, session summaries, export/import, duplicate merging, relationships/graphs, and v1.5 features like version history/rollback, templates, categories, and advanced search.
  • Simple install via npx and straightforward Claude MCP configuration; supports local or cloud Redis.
  • Emphasis on security and compliance: require authentication, TLS, access controls, auditing, and careful handling of sensitive data.
  • Designed for individuals and teams to build shared, reusable knowledge with fast performance and low operating cost.

Sentiment

Mixed to mildly skeptical. The community recognizes that LLM context persistence is a genuine problem, but the majority of substantive comments question whether a Redis-backed MCP server with 27 tools is the right solution. Most commenters favor simpler approaches — markdown files, native Claude features, or minimal tooling. The creator's responsive engagement earns goodwill, but the prevailing sentiment is that the tool is over-engineered for most developers' needs.

In Agreement

  • The problem of context loss across LLM sessions is real and universally experienced by developers using Claude daily
  • Semantic search over a large memory store is more efficient than loading entire markdown files into context, especially at scale with thousands of memories
  • Shared persistent memory between multiple Claude instances via a central Redis store is a compelling use case that markdown files cannot easily replicate
  • Workspace isolation preventing cross-project memory contamination adds value over simple file-based approaches
  • The creator's approach of using CLAUDE.md directives to guide when Claude stores and retrieves memories shows practical sophistication

Opposed

  • Exposing 27 tools to Claude unnecessarily bloats the context window when only two (Save and Search) would suffice
  • MCP servers are over-engineered for what CLI tools and markdown files already accomplish — most should be repackaged as simple CLI tools
  • Claude's native memory features (CLAUDE.md, /memory command, file-based memory system) already solve this problem without external infrastructure
  • The entire project — code, README, and HN post — appears to be AI-generated, raising questions about signal versus noise
  • Every persistent memory feature is a potential vector for prompt injection attacks, creating security risks
  • No benchmarks or comparisons against simpler solutions like text files make it hard to justify the added complexity of Redis and semantic search