dream

Consolidate session transcripts and memory files into a curated index.

70|42|Updated Mar 27, 2026

Search Tags:#workflow #logs #memory #deduplication #session #consolidation

Ready to run

npx skills add https://github.com/tranhieutt/software_development_department --skill dream-tranhieutt

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: dream
Source: https://github.com/tranhieutt/software_development_department/tree/main/.claude/skills/dream
Command: npx skills add https://github.com/tranhieutt/software_development_department --skill dream-tranhieutt

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill fixes duplicated, stale, or oversized memory files that slow future AI sessions and risk conflicting guidance.

Core Features & Use Cases

Multi-phase consolidation walks through orienting to the memory directory, gathering fresh signals, consolidating duplicates, and pruning with an approval-first workflow.
Selective focus support honors optional topic arguments so targeted knowledge stays sharp while global memory hygiene continues.
Practical governance enforces required YAML frontmatter, merges near-duplicates, flags contradictory facts, and keeps MEMORY.md under 200 lines and 25KB with a concise index.

Quick Start

Invoke the dream skill to refresh MEMORY.md and keep topic files tidy.

Frequently Asked Questions about dream

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

FAQPage Schema

How do I consolidate AI memory files to remove duplicate or stale lessons?▼

Memory consolidation merges session transcripts and topic files into a curated index, preventing duplicate or stale lessons by applying phased probing and pruning rules to keep knowledge bases tidy.

When should I clean up my MEMORY.md index after intensive conversations?▼

Memory cleanup applies when the MEMORY.md index or topic files near their limits after intensive conversations, or when drift is suspected, ensuring the index stays under 200 lines and 25KB.

What is the best way to keep AI session memories under file size limits?▼

Keeping session memories under file size limits requires pruning rules that enforce required YAML frontmatter, merge near-duplicates, and flag contradictory facts to maintain a concise index.

Can I target specific topics during memory file consolidation without a full rewrite?▼

Selective focus support honors optional topic arguments so targeted knowledge stays sharp during consolidation, while global memory hygiene continues across all session transcripts.

Does memory consolidation require user approval before editing files?▼

Memory consolidation requires user-approved file edits, applying an approval-first workflow that walks through orienting, gathering signals, consolidating duplicates, and pruning.

Why do AI sessions slow down or risk conflicting guidance over time?▼

Duplicated, stale, or oversized memory files slow future AI sessions and risk conflicting guidance, requiring consolidation to refresh the memory directory and prevent duplicate lessons.