Consolidate SQLite memory banks using forgetting curves and semantic graph connections.
npx skills add https://github.com/roy2392/scout-graph-memory --skill dream-roy2392
Please help me install this Agent Skill. Skill: dream Source: https://github.com/roy2392/scout-graph-memory/tree/main/skills/dream Command: npx skills add https://github.com/roy2392/scout-graph-memory --skill dream-roy2392
This skill prevents AI agent memory banks from becoming bloated, disorganized, and disconnected by performing an automated nightly consolidation process that mimics REM sleep.
Run the dream skill to perform a full nightly consolidation pass on the current memory bank.
High-intent search queries and answers about installing and using this skill.
To consolidate AI agent memory, you can run an automated nightly process that applies forgetting curves, merges redundant entries, and weaves semantic graph connections. This prevents your local SQLite memory store from becoming disorganized and bloated.
The best way to prune stale information from a local SQLite memory store is to use an automated memory consolidation process. It applies forgetting curves and pressure-adaptive scaling to decay stale data while maintaining a target entry count.
Yes, AI agent memory consolidation operates on local SQLite-based memory stores and requires Node.js along with the scout-graph-memory package to execute its nightly decay and weaving algorithms.
Pressure-adaptive scaling manages memory bank size by dynamically adjusting forgetting thresholds and merge aggressiveness based on current memory volume. This ensures the agent stays within its optimal performance budget.
Yes, you can establish semantic graph connections by running a memory weaving algorithm. This process links related facts during the nightly consolidation pass, ensuring better recall and graph integrity for your AI agent.
AI agents lose performance when long-term memory banks become bloated and disconnected. An automated nightly consolidation process mimics REM sleep to merge duplicates and prune noise, maintaining a compact, high-quality knowledge graph.
1thirteeng3Store and retrieve durable session memory as a Markdown knowledge graph.
paperclipaiManage persistent knowledge storage and retrieval with PARA folders and YAML facts.
an8079Train and optimize embedding models for text, images, and multimodal data.
HeadyAIOrganize and retrieve multi-tier AI agent memories with phi-decay forgetting.
