What problem does it solve? Important decisions, gotchas, and learnings from a coding session are lost when the session ends. This Skill preserves them in agentmemory's long-term storage with specific concept tags so future sessions can retrieve them via recall. ## Core Features & Use Cases - Memory Persistence: Saves insights, decisions, and facts to long-term storage through the memory_save tool with content, concepts, and file references. - Concept Tagging: Extracts 2-5 specific, lowercased concept phrases (e.g., jwt-refresh-rotation) so future recall queries actually surface the memory. - File Linking: Attaches real repo-relative or absolute file paths to memories for code-specific context. - Use Case: After deciding to drop Redis for the iii-engine state store, say "remember this decision" and the Skill saves it tagged with session-caching, iii-state-store, and dropped-redis so any later caching question retrieves it. ## Quick Start Tell the agent "remember that we rotate JWT refresh tokens on every use in auth/refresh.ts" and it will save the memory with specific concept tags and confirm the retrieval terms.
