mem0-context-loader

Searches mem0 memories and injects relevant project context before starting a task.

Updated Jun 29, 2026
One-click install
npx skills add https://github.com/nhatnguyen1122/Agent-Memory-Eval --skill mem0-context-loader-nhatnguyen1122
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: mem0-context-loader
Source: https://github.com/nhatnguyen1122/Agent-Memory-Eval/tree/main/mem0/integrations/mem0-plugin/.opencode-plugin/opencode-skills/mem0-context-loader
Command: npx skills add https://github.com/nhatnguyen1122/Agent-Memory-Eval --skill mem0-context-loader-nhatnguyen1122

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? When starting a new task or switching context, AI coding sessions lack awareness of past project decisions, coding conventions, and known pitfalls. This Skill pre-fetches relevant memories from mem0 so the session begins with the right historical context instead of rediscovering it. ## Core Features & Use Cases - Multi-angle memory search: Runs 2-4 parallel search_memories queries filtered by memory type (decision, convention, anti_pattern) plus a broad catch-all. - Deduplication and capping: Merges results by memory ID and returns at most 10 of the most relevant memories in a compact plain-text block. - Silent on empty: Outputs nothing when no relevant memories exist, avoiding noise in the session. - Use Case: A developer starts work on a payment module; the Skill loads prior architecture decisions, naming conventions, and known anti-patterns for that module before any code is written. ## Quick Start Ask the assistant to load context for the feature or files you are about to work on, for example: load context for the authentication module before I start refactoring it.

Frequently Asked Questions about mem0-context-loader

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I load project context from mem0 before starting a task?

Invoke the context loader at session start or when switching tasks. It extracts topics from your message, runs parallel search_memories queries across decision, convention, and anti-pattern filters, then outputs a compact block of up to 10 relevant memories.

What types of memories does mem0 context search retrieve?

It retrieves three filtered memory types: decisions for architecture choices, conventions for coding patterns, and anti_patterns for known pitfalls. A fourth broad query with only user_id and app_id filters acts as a catch-all for general project context.

Can the context loader modify or delete stored memories?

No, the Skill is strictly read-only. It only runs search_memories queries and never modifies or deletes memories, so your stored project history remains untouched during context loading.

Why does the context loader sometimes output nothing?

It is designed to be silent on empty results. If no relevant memories match the extracted topics, it outputs nothing rather than announcing empty context, keeping the session free of noise.

How many memories are injected into context at once?

At most 10 memories are returned, deduplicated by memory ID across all search responses. Memories already visible in the current session context are skipped to avoid redundancy.