🦊 Nexus Memory

One brain for all your agents.

Your agents forget. Your context gets lost. Your setup knowledge is scattered across chats, tools and repos.

Nexus Memory gives every MCP-compatible agent one persistent, self-hosted memory they all share.

Hermes • OpenClaw • Claude Code • Codex • Cursor • Cline • Roo Code • GitHub Copilot • Pi • Continue • Odysseus • Kilo Code …and more!

🤖 Bot Self-Install: Tell your agent: "Read AGENTS.md and install Nexus Memory." It does the rest.

👉

👉

---

---

🤖 Quick Start

Tell your agent to install it

Send this prompt to any MCP-compatible agent:

Read https://raw.githubusercontent.com/Neboy72/nexus-memory/main/AGENTS.md and follow the installation instructions.

Your agent will check prerequisites, install everything, configure the provider, and verify. Zero manual steps.

🛠️ Or install manually

git clone https://github.com/Neboy72/nexus-memory.git
cd nexus-memory
pip install -e .

Choose your embedding (auto-detected at runtime, you pick):

• 💚 Google / Vertex AI — GOOGLE_API_KEY in .env (768d)
• 💜 Jina — JINA_API_KEY in .env (1024d, best value)
• 🦙 Ollama — ollama pull nomic-embed-text
• ☁️ Voyage — VOYAGE_API_KEY in NEXUS_ENV_FILE or MCP env:-block (1024d, best quality)
• ☁️ OpenAI — OPENAI_API_KEY in NEXUS_ENV_FILE or MCP env:-block (1536d)
• 🏠 Local (default) — pip install nexus-memory[local] (sentence-transformers, no key)

Start the server:

nexus-memory

🔌 Platform Configuration

Choose your agent:

🔷 Hermes Agent

~/.hermes/config.yaml:

mcp_servers:
  nexus:
    command: nexus-memory

Restart: hermes gateway restart

🔷 OpenClaw

~/.openclaw/openclaw.json (mcp.servers.<name>.env — nested, not top-level):

{
  "mcp": {
    "servers": {
      "nexus-memory": {
        "command": "nexus-memory",
        "env": { "VOYAGE_API_KEY": "vo-your-key-here" }
      }
    }
  }
}

🔷 Claude Code

~/.claude/settings.json or .mcp.json in project root:

{
  "mcpServers": {
    "nexus": {
      "command": "python3",
      "args": ["-m", "nexus_memory.mcp_server"]
    }
  }
}

🔷 Codex CLI

~/.codex/config.toml:

[mcp_servers.nexus]
command = "python3"
args = ["-m", "nexus_memory.mcp_server"]

🔷 GitHub Copilot (VS Code)

.vscode/mcp.json in your project:

{
  "mcpServers": {
    "nexus": {
      "command": "python3",
      "args": ["-m", "nexus_memory.mcp_server"]
    }
  }
}

🔷 Cursor

Settings → Features → MCP Servers → Add:

• Name: nexus
• Command: python3
• Arguments: -m nexus_memory.mcp_server

🔷 Cline / Roo Code

MCP Server Config:

{
  "mcpServers": {
    "nexus": {
      "command": "python3",
      "args": ["-m", "nexus_memory.mcp_server"]
    }
  }
}

🔷 Kilo Code

.mcp.json in your project:

{
  "mcpServers": {
    "nexus": {
      "command": "python3",
      "args": ["-m", "nexus_memory.mcp_server"]
    }
  }
}

🔷 Pi Coding Agent

~/.pi/config.json:

{
  "mcpServers": {
    "nexus": {
      "command": "python3",
      "args": ["-m", "nexus_memory.mcp_server"]
    }
  }
}

🔷 Continue.dev

.mcp.json or ~/.continue/config.json:

{
  "mcpServers": {
    "nexus": {
      "command": "python3",
      "args": ["-m", "nexus_memory.mcp_server"]
    }
  }
}

🔷 Odysseus (PewDiePie)

Settings → MCP Management → Add Server:

• Name: nexus
• Command: python3
• Arguments: -m nexus_memory.mcp_server

🔷 Any MCP-compatible agent

Standard MCP stdio config:

{
  "mcpServers": {
    "nexus": {
      "command": "python3",
      "args": ["-m", "nexus_memory.mcp_server"]
    }
  }
}

---

🎯 MCP Tools

Tool	Description	Parameters
remember 💾	Store a memory	text (req), access_level, category, source, source_url, confidence
recall 🔍	Hybrid search (BM25 + Vector + RRF)	query (req), limit, filter_level
forget 🗑️	Delete a memory	memory_id (req)
update ✏️	Update in-place, preserve metadata	memory_id (req), text, modified_by
health ❤️	Check server status	—

Memory Categories

Category	Scope	Use Case
fact ✅	Permanent	Verified facts, decisions (default)
belief 🤔	Drift-prone	Assumptions that may change over time
session 🔄	Ephemeral	Current conversation context
rule 📏	Permanent	Operating rules, policies
preference ❤️	Permanent	User likes, dislikes, habits
temp ⏳	Temporary	Short-lived notes, TTL-managed

Access Levels 🛡️

Level	Visible to	Example
🟢 public	All agents	Project knowledge, technical info
🟡 trusted	Approved agents only	Personal preferences, habits
🔴 private	Owner only	Financial data, passwords, bills

---

✨ Features

Hybrid Retrieval 🛡️

Pure vector search is vulnerable to RAG poisoning — adversarial documents that rank high semantically but contain garbage. Nexus Memory blends BM25 + Vector + Reciprocal Rank Fusion:

Query → ┌─ BM25 Index ──────→ Keyword Rankings
         │                          │
         └─ Vector Embeddings ──→ Semantic Rankings
                                        │
                               RRF Fusion ───→ Combined Rankings

Method	Strengths	Weaknesses
BM25 🔤	Keyword-exact, poison-resistant	Misses semantics
Vector 🧠	Semantic matching, fuzzy queries	Vulnerable to poisoning
Hybrid (RRF) 🏆	Best of both	—

Source-Tier Boosting 🏷️

Tier	Sources	Boost
🟢 Tier 1	Agent, user, official docs	1.2×
🟡 Tier 2	Curated external	1.0×
🔴 Tier 3	Uncurated / unknown	0.8×

MemoryCategory Enum 🏷️

Six scopes from Agentic Design Patterns (Ch8): fact, belief, session, rule, preference, temp. Every memory knows its purpose.

Provenance Tracking 📎

Every memory carries its origin: source_url, confidence (0.0–1.0), modified_by, timestamps. Full audit trail from creation to today.

Guardrails 🛡️

Content-length warnings for entries >5,000 chars. PII detection hints for emails and phone numbers in non-private entries.

Fact Lifecycle Model 🧬

Append-only state machine: pending → canonical | deprecated | rolled_back. Every revision is versioned with fact_id, version_id, content_hash, supersedes, and mandatory decision_event. No silent overwrites. No zombie facts.

Staging + Rollback 🔄

Operation	What it does
create_pending()	Stage new facts for review
promote()	Promote staged → canonical
deprecate()	Mark canonical as deprecated
rollback()	Restore previous canonical version

Auto-Discovery + Graph Analytics 🔄

Zero-token relation discovery between canonical facts via Qdrant (O(n·k)) + heuristic classification. Graph analytics: hub scores, isolation scores, knowledge gaps, connected components. Facts connect themselves — no manual edges needed.

Skill Export 🎯

export_skill() searches canonical facts → clusters into Steps/Pitfalls/Prerequisites/Verification → generates complete SKILL.md. Turn learned facts into reusable agent skills.

Belief Drift Detection 🔍

Score	Status
🟢 < 1	Healthy
🟡 1–3	Attention needed
🔴 > 3	Action required

Detects stale entries, old patterns ("X running as fallback" — but X was replaced), age thresholds. Weighted 0–10 scoring.

---

📊 vs Other Memory Solutions

Feature	Nexus Memory 🦊	Walrus Memory 🦭	mem0	Honcho	agentmemory	Holographic
🔍 Semantic search	✅ local or cloud	✅ via API	✅ Cloud	✅ pgvector	✅ Gemini	✅ HRR algebra
🔀 Hybrid retrieval	✅ BM25 + Vector + RRF	❌	✅ Multi-signal	❌	❌	❌
🩺 Drift detection	✅ Scored 0–10	❌	❌ *	❌	❌	❌
🛡️ Anti-poisoning	✅ Source tiers	❌	❌	❌	❌	❌
🔗 Multi-Level Provenance	✅ Source + Corroboration + Dep.	✅ On-chain	❌	❌	❌	❌
🏷️ MemoryCategory Enum	✅ 6 scopes	❌	❌	❌	❌	❌
🧬 Fact Lifecycle	✅ Append-only	❌	❌	❌	❌	❌
🔄 Staging + Rollback	✅ Promote/Deprecate/Rollback	❌	❌	❌	❌	❌
🎯 Skill Export	✅ Facts → SKILL.md	❌	❌	❌	❌	❌
🔗 SkillGraph	✅ 5 relation types, BFS/DFS	❌	❌	❌	❌	❌
🔄 Auto-Discovery	✅ 0 token cost	❌	❌	❌	❌	❌
📊 Graph Analytics	✅ Hub scores, gaps	❌	❌	❌	❌	❌
🚀 Graph Boost	✅ Search ranking boost	❌	❌	❌	❌	❌
🛡️ Access Control	✅ public/trusted/private	✅ Permissions	❌	❌	❌	❌
🏠 Self-hosted	✅ Your machine	❌ Blockchain	❌ Cloud	❌ Cloud	❌ Cloud	✅ Local
💰 Cost	🆓 Free	WAL token	Subscription	Subscription	API costs	Free
📦 Code size	~9.6K Python	Managed service	Managed service	Managed service	~50K TS	~1.5K Python
⏱️ Setup time	1 command	Signup + SDK	API key + signup	Postgres + pgvector	30+ min + OAuth	1 command

*Mem0 lists staleness as an "open problem" in their 2026 report but does not ship a solution.

Nexus Memory is the only solution with hybrid retrieval, drift detection, provenance, fact lifecycle, staging/rollback, auto-discovery, graph analytics, skill export, memory categories, and access control — all self-hosted, all in one package.

---

🧩 Embedding Providers

One server. Multiple backends. Same API.

Provider	Type	Setup	Dims	Quality
Voyage ☁️	Cloud API	VOYAGE_API_KEY in .env	1024	⭐ Best
OpenAI ☁️	Cloud API	OPENAI_API_KEY in .env	1536	⭐ Great
Ollama 🦙	Local	ollama pull nomic-embed-text	768	Better
sentence-transformers 🏠	Local	pip install sentence-transformers	384	Good ✅ (default)

---

🔧 Troubleshooting

Symptom	Check	Fix
mcp_nexus_* tools missing	grep 'nexus' ~/.hermes/logs/agent.log	Gateway restart
Qdrant not running	curl http://127.0.0.1:6333/healthz	brew services start qdrant
Hybrid search missing	pip list | grep bm25s	pip install bm25s
Voyage embedding fails	echo $VOYAGE_API_KEY	Set in ~/.hermes/.env
ModuleNotFoundError	Check PYTHONPATH	Set PYTHONPATH=/path/to/nexus-memory

---

🧪 Tests

pytest tests/ -v   # 224 tests ✅

---

📋 Requirements

• Python 3.11+
• Qdrant v1.12+ running on localhost:6333
• One embedding provider (auto-detected):
  • 💚 Google / Vertex AI — GOOGLE_API_KEY in .env (768d)
• 💜 Jina — JINA_API_KEY in .env (1024d, best value)
• 🦙 Ollama — ollama pull nomic-embed-text
  • ☁️ Voyage — VOYAGE_API_KEY in .env
  • ☁️ OpenAI — OPENAI_API_KEY in .env
  • 🏠 Local — pip install sentence-transformers
• Optional: bm25s for hybrid search

---

📜 License

MIT — use it, modify it, ship it.

---

⭐️ Found it useful? Give it a star on GitHub — it helps others find it!

_{Built by Nebo · June 2026 · v0.2.0 — One memory for all your agents}