skills.rest
kaifashraff
kaifashraffCOMMUNITY
Explore all skills by this author →
GitHub

fluid-memory

Store, recall, and archive conversation memories with decay and semantic search.

Updated Mar 18, 2026
Search Tags:#recall#memory#vector#openclaw#chromadb#forget#fluid-memory
One-click install
npx skills add https://github.com/kaifashraff/jarvis-research --skill fluid-memory-kaifashraff
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: fluid-memory
Source: https://github.com/kaifashraff/jarvis-research/tree/main/skills/fluid-memory
Command: npx skills add https://github.com/kaifashraff/jarvis-research --skill fluid-memory-kaifashraff

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Fluid Memory manages long-running conversations by selectively storing important memories, surface recalling when needed, and fading or archiving less relevant content to keep context lightweight.

Core Features & Use Cases

  • Dynamic forgetting and sentiment-aware archiving to balance recall relevance.
  • Semantic recall via a vector store (ChromaDB) to surface relevant memories.
  • Incremental summarization and hooks to integrate with OpenClaw workflows.
  • Privacy-conscious by design with local storage and configurable thresholds.

Quick Start

Install the Fluid Memory Skill and use remember to store memories and recall to retrieve them.

Frequently Asked Questions about fluid-memory

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

FAQPage Schema
How do I manage long-context memory in conversational AI without exceeding token limits?▼

Long-context memory management uses semantic search over vector stores like ChromaDB to store and recall important conversations, while dynamically forgetting low-value items to keep context lightweight.

How does dynamic forgetting work for conversation memory?▼

Dynamic forgetting uses a fluid decay and reinforcement model to fade or archive less relevant content, balancing recall relevance by sentiment-aware archiving and configurable thresholds for outdated items.

Can I integrate semantic recall with OpenClaw workflows?▼

Yes, semantic recall integrates with OpenClaw workflows through memory hooks and optional incremental summarization, allowing you to surface relevant vector store memories directly within your automated pipelines.

Do I need a vector database to use fluid memory recall?▼

A vector database like ChromaDB powers semantic recall, but fluid memory includes a fallback for non-vector environments, allowing basic conversation storage and retrieval without a dedicated vector store.

How do I store and retrieve conversation memories locally?▼

You store and retrieve conversation memories locally by applying privacy-conscious local storage, using commands to remember new inputs and recall relevant past context via semantic search.

What are the limitations of using a fluid decay model for conversation archiving?▼

The fluid decay model requires careful threshold configuration to prevent premature forgetting of valuable context, as sentiment-aware archiving may deprioritize technically important but low-sentiment memories.

Related Skills

Explore More Software Engineering Skills→
Rawgrowth-ConsultingRawgrowth-Consulting

chroma

Store embeddings with metadata in a local Chroma vector database for similarity search.

Official
Intermediate
giosuetedeschi-specgiosuetedeschi-spec

context-window-management

Summarize, trim, and route LLM context to prevent token exhaustion.

Community
Advanced
jholhewresjholhewres
14

memory

Store and retrieve long-term memories with semantic search and category tagging.

Official
Intermediate
winsorllcwinsorllc

memory

Store, recall, and forget key-value memories across sessions using JSON.

Community
Intermediate