Auto-discovered marketplace from adverant/adverant-nexus-cowork-plugin
npx claudepluginhub adverant/adverant-nexus-cowork-pluginGive Claude persistent memory, a skills marketplace, and multi-agent superpowers. Claude forgets everything between conversations — Nexus fixes that. Store memories in a GraphRAG knowledge graph that persists forever, discover and invoke 50+ reusable AI skills, synthesize new workflows by combining existing ones, manage documents with semantic search, orchestrate multiple AI agents for complex tasks, access all marketplace plugin tools through a unified dispatcher, and enforce global session directives (memory protocol, engineering standards, review gates). 9 tools, 7 skills, 3 commands — one plugin.
Share bugs, ideas, or general feedback.
Give Claude persistent memory, a skills marketplace, and multi-agent superpowers.
Every time you start a new conversation with Claude, it forgets everything. Your preferences, your project context, the skills you taught it, the documents you shared — gone. You start over. Every. Single. Time.
You've built workflows, refined prompts, curated knowledge — but none of it persists. Claude is powerful, but without memory, it's like hiring a brilliant contractor who shows up with amnesia every morning.
What if Claude could remember everything, access a marketplace of reusable skills, and coordinate multiple agents — all from a single plugin?
Open Claude Desktop and paste the GitHub repo URL into Browse Plugins:
adverant/adverant-nexus-cowork-plugin
That's it. Three steps, thirty seconds:
adverant/adverant-nexus-cowork-plugin → Sync → InstallNo config files. No terminal commands. No API keys to manage.
┌─────────────────────────┐
│ Adverant Nexus Plugin │
│ │
│ 8 Tools · 6 Skills │
│ 3 Commands · 1 Agent │
└────────────┬────────────┘
│
┌──────────────────────┼──────────────────────┐
│ │ │
┌─────────▼─────────┐ ┌────────▼────────┐ ┌─────────▼─────────┐
│ Persistent Memory │ │ Skills Engine │ │ Multi-Agent │
│ │ │ │ │ Orchestration │
│ • GraphRAG store │ │ • 50+ skills │ │ │
│ • Knowledge graph │ │ • Discover │ │ • Collaborate │
│ • Entity tracking │ │ • Invoke │ │ • Compete │
│ • Semantic search │ │ • Synthesize │ │ • Analyze │
│ • Document mgmt │ │ • Auto-sync │ │ • Synthesize │
└───────────────────┘ └─────────────────┘ └───────────────────┘
Claude normally loses all context between conversations. With Nexus, everything you tell it persists in a GraphRAG knowledge graph — relationships, facts, preferences, project context. Ask Claude to "remember that our API uses JWT auth" and it's stored permanently. Next week, next month — it still knows.
You: "Remember that our production database is on us-east-1
and we use Prisma as our ORM."
Claude: Stored in your knowledge graph.
--- Two weeks later, new conversation ---
You: "What ORM do we use?"
Claude: You use Prisma as your ORM, with your production
database hosted on us-east-1.
Stop rebuilding the same workflows from scratch. The Skills Engine gives Claude access to 50+ reusable, tested skills — from code review to patent creation to SEO optimization. Discover skills by describing what you need, invoke them instantly, or synthesize entirely new skills by combining existing ones.
You: "Find me a skill for code review"
Claude: Found 3 matching skills:
1. code-review (quality: 94%) — Multi-perspective security,
logic, and integration analysis
2. sp-requesting-code-review — Verify work meets requirements
3. sp-receiving-code-review — Technical rigor for feedback
You: "Combine code-review with the security analysis skill"
Claude: Synthesized new skill: secure-code-review
Strategy: chain (code-review → security-analysis)
Quality score: 96%
Some tasks are too complex for a single pass. Nexus lets Claude coordinate multiple AI agents that can collaborate, compete, or specialize — then synthesize the best result.
You: "Have 3 agents analyze this architecture proposal —
one for security, one for performance, one for cost."
Claude: Orchestrating 3 specialized agents...