Skip to content

TeamADAPT/Minimax-Mini-Agent

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

42 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

NovaOps — ADAPT Autonomous Agent Platform

TeamADAPT | Bleeding-edge AIML R&D | Charter | Protocols


What This Is

NovaOps is the infrastructure layer for deploying, orchestrating, and maintaining autonomous Nova agents within the TeamADAPT ecosystem. It provides identity continuity, tool management, cross-agent communication, and the control plane through which agents and human operators interact.


System Requirements

  • Linux (systemd)
  • Python 3.10+ (system-wide, no venv)
  • Node.js 20+ (for Paperclip control plane)
  • NATS Server (localhost:18020)
  • Qdrant (localhost:6333), Redis, MongoDB, Neo4j, ClickHouse
  • Temporal (localhost:8233)
  • Local LLM endpoint (localhost:8888/v1)
  • No Docker. All services managed via systemd.

Directory Structure

novaops/
├── controlplane/           Paperclip (AI company OS) + Pipecat
│   └── paperclip/          13 ADAPT autonomous plugins (see below)
├── nova_framework/         Core agent runtime — NATS event hub, bridge service
├── novas/                  Named Nova agent instances (nova_001–010 + specialists)
├── toolops/                70+ MCP servers, API, DB, voice, UI, memory layers
├── novamem/                Memory backend — Qdrant + MongoDB + Redis + Weaviate
├── novamon/                Observability — LangFuse + LangSmith
├── polyglot-agents/        Multi-language agent support
├── frameworks/             15 framework integrations (autogen, crewai, camel, etc.)
├── continuity/             Cross-session continuity, comms, novathreads
├── x1-wasm_rust/           Rust/WASM component
├── core_infra/             NATS client
├── nova_core/              Nova core runtime
├── nova_infra/             Secrets + realtime comms (restricted access)
├── novainfra/              Nova infrastructure layer
├── protocols/              Governance: charter, protocols, onboarding
├── ops/                    Task workflow + operational logs
│   ├── to_do/
│   ├── in_progress/
│   └── completed/
├── mini_agent/             Mini Agent — base agent implementation (MiniMax M2 / Anthropic-compatible)
├── scripts/                Utility scripts
├── tests/                  Test suite
└── docs/                   Development and production guides

Paperclip Plugins (Control Plane)

All 13 ADAPT autonomous plugins run inside paperclip.service (systemd):

Plugin Function
Command Center CEO dashboard — agent chat, widgets, mobile controls
Conductor Agent instruction editor, AI-assisted improvement, revision history
Nexus Multi-agent group chat, collaboration rooms, debate arena
Bridge GitHub / Slack / Discord / Jira — bidirectional sync
Chain Temporal workflow orchestration — durable agent pipelines
Incident Commander War room — declare incidents, coordinate response, post-mortem
Triage Semantic issue intake, LLM classification, priority routing
Ops Dead Man's Switch, SLA enforcer, board reports
Observer NATS stream monitor + ClickHouse metrics dashboard
OKR Neo4j-backed objectives and key results, agent alignment
Memory Semantic decision log, Qdrant vector search, RAG retrieval
Vault Prompt library — store, tag, version, semantic search
Chat LLM-powered brainstorming interface

Service Management

All services run via systemd. Key units:

systemctl status paperclip          # Paperclip control plane (port 3100)
systemctl status nats               # NATS event bus (port 18020)

Task Workflow

ops/to_do/  →  ops/in_progress/  →  ops/completed/

All actions logged in ops/operations_history.md and ops/decisions.log — timestamped, signed by agent.


Ops Discipline

Every operational action must be logged:

  • ops/operations_history.md — reverse-chronological, newest first
  • ops/decisions.log — all decisions with rationale

Entry format:

## YYYY-MM-DD HH:MM:SS — AGENT_ID
Description of action or decision.

Governance


License

MIT

About

A minimal yet professional single agent demo project that showcases the core execution pipeline and production-grade features of agents.

Resources

License

Code of conduct

Contributing

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages

  • Python 87.5%
  • HTML 4.8%
  • JavaScript 4.1%
  • PLpgSQL 1.7%
  • Shell 1.5%
  • PowerShell 0.4%