Add multimodal lifelong memory to MCP and Python agents with SimpleMem
Use SimpleMem to store, compress, index, and retrieve text or multimodal memories for agents through MCP or Python integrations.
npx skills add agentskillexchange/skills --skill add-multimodal-lifelong-memory-to-mcp-and-python-agents-with-simplemem
Use SimpleMem when an operator needs an agent to retain and retrieve long-lived dialogue facts, user preferences, or multimodal observations without replaying full transcripts. The workflow is to install the simplemem Python package or connect the hosted MCP server, add dialogue, text, image, audio, or video memories through the documented API, finalize or index the memory store, and query relevant memories before an agent responds or plans. Invoke this instead of using the product normally when the job is a bounded memory runbook for an MCP client or Python-based agent, not generic vector search or research-paper reproduction. The boundary is storing, compressing, indexing, and retrieving scoped memories for future agent context; it is not a broad framework, model, or database listing.
What this skill actually does
Inputs and prerequisites: SimpleMem package or MCP server, an MCP-capable client or Python agent runtime, memory inputs such as dialogues/text/media files, and any configured model or embedding dependencies required by the selected backend.
Setup notes: Install the simplemem package from PyPI for Python workflows or configure the documented MCP server for text-memory workflows, add scoped memories with the provided APIs, finalize/index the memory set, and query only relevant memories during agent runs.
Source and verification boundary: use https://github.com/aiming-lab/SimpleMem and its MCP/package documentation as the canonical reference before running the workflow; keep stored memories, retrieval parameters, and query outputs reviewable.
Framework fit: publish this as an MCP workflow because upstream explicitly supports MCP clients for text memory, while still documenting Python integration for fuller multimodal memory.