repoGitHubTrust 82 · PrimaryPublished 1mo agoLive · 1mo ago
nossa-y/activity-frames
Episodic memory for AI agents. Records your screen locally, compiles it into structured activity frames, serves them over MCP. No cloud, no LLM in the loop.
Lineage graph
Paper → model → repo connections mined from source citations (Tier-1 exact match).
Why these links exist
Every edge carries a method, confidence, and the source snippet that justified it — so bad links are debuggable.
- PossiblePossibly related (embedding) · 56%We're building agents that can read millions of documents, but still forget a video they watched yesterday. →
- PossiblePossibly related (embedding) · 52%MRMS: A Multi-Resolution Memory Substrate for Long-Lived AI Agents →
- PossiblePossibly related (embedding) · 49%AgenticSTS: A Bounded-Memory Testbed for Long-Horizon LLM Agents →
- PossiblePossibly related (embedding) · 49%Always-OnAgents:A Survey of Persistent Memory, State, and Governance in LLMAgents →
- PossiblePossibly related (embedding) · 52%Show HN: Screen memory without screenshots, just text to Markdown →
- PossiblePossibly related (embedding) · 55%LightMem-Ego: Your AI Memory for Everyday Life →
Covers
Implements
Covers (incoming)
Implements (incoming)
Related across the graph
paperAlways-OnAgents:A Survey of Persistent Memory, State, and Governance in LLMAgentsnewsWe're building agents that can read millions of documents, but still forget a video they watched yesterday.paperMRMS: A Multi-Resolution Memory Substrate for Long-Lived AI AgentspaperAgenticSTS: A Bounded-Memory Testbed for Long-Horizon LLM AgentspaperLightMem-Ego: Your AI Memory for Everyday LifenewsShow HN: Screen memory without screenshots, just text to Markdown