Back to browse
GitHub Repository

Specialist AI engineering team for Claude Code. 8 agents with HITL checkpoints and /agency-run orchestrator. No platform. No login. MIT.

13 starsShell

Navox Agents – 8 AI Agents for Claude Code with HITL Checkpoints

by nahrin·Apr 17, 2026·1 point·1 comment

AI Analysis

●●SolidNiche GemShip It

Multi-agent orchestration for Claude Code CLI with human checkpoints instead of full autonomy.

Strengths
  • Human-in-the-loop checkpoints prevent runaway agents from deleting production code.
  • Zero platform lock-in; runs locally via Claude Code plugin system.
  • Demonstrated output includes complex projects like PipeWar with passing tests.
Weaknesses
  • Tightly coupled to Claude Code CLI; useless if you switch editors.
  • Multi-agent prompt chains often brittle compared to compiled orchestration logic.
Target Audience

Developers using Claude Code CLI who want automated workflows

Similar To

Cline · Aider · Cursor

Similar Projects

AI/ML●●Solid

Sovereign – Multi-agent OS with GraphRAG memory and HITL checkpoints

Runtime pause-and-resume checkpoints during execution plus per-agent escalation policies are a rare, practical safety feature — not just a pre-flight approval checkbox. The hybrid memory that scores vector similarity, BM25 keyword relevance, and entity/relation overlap together is a smart middle path between heavyweight graph DBs and plain vector stores. There's real engineering here (Docker compose, 55/55 tests, telemetry endpoints), though I'd like to see benchmarks and how the in-process GraphRAG scales versus external vector/graph services.

WizardryNiche Gem
borhensaidi
103mo ago