Consolidate Claude Code memory files by merging, pruning, and cross-referencing.
npx skills add https://github.com/moondrop-ro/claude-skills --skill memory-dream
Please help me install this Agent Skill. Skill: memory-dream Source: https://github.com/moondrop-ro/claude-skills/tree/main/skills/memory-dream Command: npx skills add https://github.com/moondrop-ro/claude-skills --skill memory-dream
Memory fragmentation across Claude Code sessions is reduced by automatically discovering, merging near-duplicate memories, pruning stale entries, and cross-referencing with CLAUDE.md to keep per-project memory stores coherent.
Run memory-dream to scan the current project and ancestors, then apply automatic memory consolidation.
High-intent search queries and answers about installing and using this skill.
Memory consolidation for Claude Code is automated by discovering, merging near-duplicate memories, pruning stale entries, and organizing files across project hierarchies. It cross-references CLAUDE.md to maintain consistency across per-project and ancestor memories.
Memory deduplication identifies and merges near-duplicate memory files to reduce fragmentation and confusion. You need it for multi-project codebases when per-project memories, ancestor memories, and cross-references become fragmented across sessions.
Yes, you can verify memories against the codebase to ensure alignment using auto mode. Auto mode applies changes automatically and summarizes results after each run, keeping per-project memory stores coherent with the actual codebase.
Yes, it handles per-project memories and ancestor memories across project hierarchies. It discovers memory files throughout the hierarchy, applies staleness detection and deduplication, and maintains cross-references against CLAUDE.md for consistency.
The best way to prune stale memories is running an automated consolidation scan that detects staleness, removes outdated entries, and clusters related memories into coherent topic files. Plan mode previews changes while auto mode applies them directly.
Memory fragmentation causes confusion because near-duplicate memories accumulate across sessions and project hierarchies without deduplication. Stale entries persist, and per-project memories drift out of sync with CLAUDE.md and the actual codebase.
slbugCapture and store lessons from bug fixes with duplicate detection and consent.
wallterAudit TRW memory health and generate a structured report with recommendations.
bugrabilgeScan project files and store context, patterns, and decisions in markdown memory files.
johnatas-henriqueAggregate sprint metrics and flag stale tasks from plan and story files.
