From 1ed8eb339d0739a28a5b9c443b681dfa7156ea21 Mon Sep 17 00:00:00 2001 From: EverestAn Date: Sun, 23 Aug 2026 14:07:56 +1000 Subject: [PATCH 1/4] Add Awareness to Memory Systems --- README.md | 1 + 1 file changed, 1 insertion(+) diff --git a/README.md b/README.md index 9e4a54a..17e3b75 100644 --- a/README.md +++ b/README.md @@ -74,6 +74,7 @@ Vector, graph, episodic, and hybrid memory architectures for persistent agent co - [**memgraph**](https://github.com/memgraph/memgraph) - High-performance open-source in-memory graph database for GraphRAG, AI memory, agentic AI, and real-time graph analytics. Cypher-compatible, built in C++. by [@memgraph](https://github.com/memgraph) (4,347 stars) - [**Acontext**](https://github.com/memodb-io/Acontext) - Open-source skill memory layer for AI agents. Automatically captures learnings from agent runs and stores them as reusable skill files. by [@memodb-io](https://github.com/memodb-io) (3,673 stars) - [**ReMe**](https://github.com/agentscope-ai/ReMe) - Memory management kit for agents. File-based and vector-based memory systems. SOTA on LoCoMo and HaluMem benchmarks. by [@agentscope-ai](https://github.com/agentscope-ai) (3,325 stars) +- [**Awareness**](https://github.com/everest-an/Awareness) - Universal external memory for AI agents: hybrid retrieval (Qdrant+pgvector, BM25+vector RRF), conflict detection, knowledge cards, bi-temporal graph memory. 95.6% recall@5 on LongMemEval_S locally on an M1 with zero LLM calls. Local-first and free forever; works with Claude Code, Cursor, Windsurf and 12+ IDEs via MCP; Google Drive / Notion / Feishu OAuth import; ERC-8350 verifiable memory state. Benchmarks: https://awareness.market/benchmarks by [@everest-an](https://github.com/everest-an) - [**MemMachine**](https://github.com/MemMachine/MemMachine) - Universal memory layer for AI agents. Episodic (graph-based), profile (SQL), and working memory with scalable storage and retrieval. by [@MemMachine](https://github.com/MemMachine) (3,178 stars) - [**datachain**](https://github.com/datachain-ai/datachain) - Operational data context layer for AI agents providing typed and versioned datasets over multimodal content. by [@datachain-ai](https://github.com/datachain-ai) (2,807 stars) - [**nocturne_memory**](https://github.com/Dataojitori/nocturne_memory) - Lightweight, rollbackable Long-Term Memory Server for MCP Agents with graph-like structured memory. by [@Dataojitori](https://github.com/Dataojitori) (1,320 stars) From 68a25855cda5afb965dff7d383e7b0094855c9cf Mon Sep 17 00:00:00 2001 From: Everest Date: Fri, 28 Aug 2026 07:27:53 +1000 Subject: [PATCH 2/4] Fix link: point at the public repo, and describe the local build accurately --- README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/README.md b/README.md index 17e3b75..eb64b94 100644 --- a/README.md +++ b/README.md @@ -74,7 +74,7 @@ Vector, graph, episodic, and hybrid memory architectures for persistent agent co - [**memgraph**](https://github.com/memgraph/memgraph) - High-performance open-source in-memory graph database for GraphRAG, AI memory, agentic AI, and real-time graph analytics. Cypher-compatible, built in C++. by [@memgraph](https://github.com/memgraph) (4,347 stars) - [**Acontext**](https://github.com/memodb-io/Acontext) - Open-source skill memory layer for AI agents. Automatically captures learnings from agent runs and stores them as reusable skill files. by [@memodb-io](https://github.com/memodb-io) (3,673 stars) - [**ReMe**](https://github.com/agentscope-ai/ReMe) - Memory management kit for agents. File-based and vector-based memory systems. SOTA on LoCoMo and HaluMem benchmarks. by [@agentscope-ai](https://github.com/agentscope-ai) (3,325 stars) -- [**Awareness**](https://github.com/everest-an/Awareness) - Universal external memory for AI agents: hybrid retrieval (Qdrant+pgvector, BM25+vector RRF), conflict detection, knowledge cards, bi-temporal graph memory. 95.6% recall@5 on LongMemEval_S locally on an M1 with zero LLM calls. Local-first and free forever; works with Claude Code, Cursor, Windsurf and 12+ IDEs via MCP; Google Drive / Notion / Feishu OAuth import; ERC-8350 verifiable memory state. Benchmarks: https://awareness.market/benchmarks by [@everest-an](https://github.com/everest-an) +- [**Awareness**](https://github.com/everest-an/Awareness-Market) - Local-first persistent memory for coding agents. 96.0% Recall@5 on LongMemEval (ICLR 2025, 500 questions) with zero LLM calls at retrieval — hybrid BM25 + vector RRF over SQLite FTS5. Runs offline, no account or API key. `npx @awareness.market/local start` - [**MemMachine**](https://github.com/MemMachine/MemMachine) - Universal memory layer for AI agents. Episodic (graph-based), profile (SQL), and working memory with scalable storage and retrieval. by [@MemMachine](https://github.com/MemMachine) (3,178 stars) - [**datachain**](https://github.com/datachain-ai/datachain) - Operational data context layer for AI agents providing typed and versioned datasets over multimodal content. by [@datachain-ai](https://github.com/datachain-ai) (2,807 stars) - [**nocturne_memory**](https://github.com/Dataojitori/nocturne_memory) - Lightweight, rollbackable Long-Term Memory Server for MCP Agents with graph-like structured memory. by [@Dataojitori](https://github.com/Dataojitori) (1,320 stars) From 83b9527fd4ab5e494fc4d5614ac851a3eae55903 Mon Sep 17 00:00:00 2001 From: EverestAn Date: Tue, 6 Oct 2026 13:40:36 +1100 Subject: [PATCH 3/4] Add Awareness Local (memory): local-first MCP memory for AI coding agents --- README.md | 2 +- data/projects.json | 16 ++++++++++++++++ 2 files changed, 17 insertions(+), 1 deletion(-) diff --git a/README.md b/README.md index eb64b94..537c55d 100644 --- a/README.md +++ b/README.md @@ -74,7 +74,6 @@ Vector, graph, episodic, and hybrid memory architectures for persistent agent co - [**memgraph**](https://github.com/memgraph/memgraph) - High-performance open-source in-memory graph database for GraphRAG, AI memory, agentic AI, and real-time graph analytics. Cypher-compatible, built in C++. by [@memgraph](https://github.com/memgraph) (4,347 stars) - [**Acontext**](https://github.com/memodb-io/Acontext) - Open-source skill memory layer for AI agents. Automatically captures learnings from agent runs and stores them as reusable skill files. by [@memodb-io](https://github.com/memodb-io) (3,673 stars) - [**ReMe**](https://github.com/agentscope-ai/ReMe) - Memory management kit for agents. File-based and vector-based memory systems. SOTA on LoCoMo and HaluMem benchmarks. by [@agentscope-ai](https://github.com/agentscope-ai) (3,325 stars) -- [**Awareness**](https://github.com/everest-an/Awareness-Market) - Local-first persistent memory for coding agents. 96.0% Recall@5 on LongMemEval (ICLR 2025, 500 questions) with zero LLM calls at retrieval — hybrid BM25 + vector RRF over SQLite FTS5. Runs offline, no account or API key. `npx @awareness.market/local start` - [**MemMachine**](https://github.com/MemMachine/MemMachine) - Universal memory layer for AI agents. Episodic (graph-based), profile (SQL), and working memory with scalable storage and retrieval. by [@MemMachine](https://github.com/MemMachine) (3,178 stars) - [**datachain**](https://github.com/datachain-ai/datachain) - Operational data context layer for AI agents providing typed and versioned datasets over multimodal content. by [@datachain-ai](https://github.com/datachain-ai) (2,807 stars) - [**nocturne_memory**](https://github.com/Dataojitori/nocturne_memory) - Lightweight, rollbackable Long-Term Memory Server for MCP Agents with graph-like structured memory. by [@Dataojitori](https://github.com/Dataojitori) (1,320 stars) @@ -83,6 +82,7 @@ Vector, graph, episodic, and hybrid memory architectures for persistent agent co - [**Awesome-Agent-Memory**](https://github.com/TeleAI-UAGI/Awesome-Agent-Memory) - Curated systems, benchmarks, and papers on memory for LLMs/MLLMs -- long-term context, retrieval, and reasoning. by [@TeleAI-UAGI](https://github.com/TeleAI-UAGI) (585 stars) - [**MemSkill**](https://github.com/ViktorAxelsen/MemSkill) - Learning and evolving memory skills for self-evolving agents. Meta-memory that determines what to extract, remember, and forget. by [@ViktorAxelsen](https://github.com/ViktorAxelsen) (562 stars) - [**TeleMem**](https://github.com/TeleAI-UAGI/telemem) - High-performance drop-in Mem0 replacement. 19% higher accuracy, 43% fewer tokens, and 2.1x speedup via narrative dynamic extraction. by [@TeleAI-UAGI](https://github.com/TeleAI-UAGI) (484 stars) +- [**Awareness Local**](https://github.com/everest-an/Awareness-Market) - Local-first MCP memory server for AI coding agents. Persistent cross-session memory built from Markdown + SQLite FTS5 + local embeddings (hybrid RRF); 96.0% Recall@5 on LongMemEval with zero LLM calls at retrieval time and fully offline operation. by [@everest-an](https://github.com/everest-an) (12 stars) ## Agent-to-Agent Protocols diff --git a/data/projects.json b/data/projects.json index d10bd20..3deb78f 100644 --- a/data/projects.json +++ b/data/projects.json @@ -1438,5 +1438,21 @@ "maintainer": "google-agentic-commerce", "tags": [], "stars": 553 + }, + { + "name": "Awareness Local", + "repo": "everest-an/Awareness-Market", + "description": "Local-first MCP memory server for AI coding agents. Persistent cross-session memory built from Markdown + SQLite FTS5 + local embeddings (hybrid RRF); 96.0% Recall@5 on LongMemEval with zero LLM calls at retrieval time and fully offline operation.", + "category": "memory", + "maintainer": "everest-an", + "tags": [ + "local-first", + "mcp", + "offline", + "sqlite", + "agent-memory" + ], + "stars": 12, + "paper": "https://arxiv.org/abs/2410.10813" } ] From 9dc5b91b35e50be4cf3b9066765821b88a69a69b Mon Sep 17 00:00:00 2001 From: autogame-17 <17@evomap.ai> Date: Tue, 20 Oct 2026 21:51:30 +0800 Subject: [PATCH 4/4] fix: correct Awareness Local catalog metadata Limit tags to the schema range and remove unsupported benchmark and paper assertions. Co-Authored-By: Claude Sonnet 5.5 --- data/projects.json | 7 ++----- 1 file changed, 2 insertions(+), 5 deletions(-) diff --git a/data/projects.json b/data/projects.json index 3deb78f..3a9d6f5 100644 --- a/data/projects.json +++ b/data/projects.json @@ -1442,17 +1442,14 @@ { "name": "Awareness Local", "repo": "everest-an/Awareness-Market", - "description": "Local-first MCP memory server for AI coding agents. Persistent cross-session memory built from Markdown + SQLite FTS5 + local embeddings (hybrid RRF); 96.0% Recall@5 on LongMemEval with zero LLM calls at retrieval time and fully offline operation.", + "description": "Local-first MCP memory server for AI coding agents with Markdown-backed cross-session memory, SQLite FTS5, and local embedding retrieval.", "category": "memory", "maintainer": "everest-an", "tags": [ "local-first", "mcp", - "offline", - "sqlite", "agent-memory" ], - "stars": 12, - "paper": "https://arxiv.org/abs/2410.10813" + "stars": 12 } ]