OpenNexum
Contract-driven coding agent orchestration via OpenClaw ACP.
When to use
- - Coordinating multiple AI coding agents (Codex + Claude)
- Running generator/evaluator pairs with automatic retry and escalation
- Parallel task execution with independent ACP sessions
- Tracking task progress with OpenClaw notifications
Architecture
Dispatch (dual-path)
- 1. Webhook (real-time):
nexum callback → POST /hooks/agent → orchestrator wakes up immediately - Dispatch Queue (fallback):
nexum callback → writes nexum/dispatch-queue.jsonl → heartbeat processes within 10min
Cross-review
- - codex-gen → claude-eval, claude-gen → codex-eval
Auto-routing
- -
generator: auto in Contract YAML → system selects agent by task type
Batch progress
- -
nexum status shows current batch progress + overall progress
Agent Naming: <model>-<role>-<number>
- - codex-gen-01~03: backend/API code
- claude-gen-01~02: user-facing WebUI/docs
- codex-eval-01 / claude-eval-01: cross-review
- claude-plan-01: architecture (opus)
Key CLI Commands
CODEBLOCK0
Contract YAML
CODEBLOCK1
Callback Protocol (injected into AGENTS.md via nexum init)
INLINECODE7 writes the callback protocol into AGENTS.md as the source of truth. If a project only has CLAUDE.md, Nexum will seed AGENTS.md from it and update AGENTS.md going forward.
After completing a task, run:
CODEBLOCK2
Git Convention
- - Push directly to
main, revert if needed - English Conventional Commits: INLINECODE13
OpenNexum
通过OpenClaw ACP实现的契约驱动编码代理编排。
使用场景
- - 协调多个AI编码代理(Codex + Claude)
- 运行生成器/评估器配对,支持自动重试和升级
- 使用独立ACP会话进行并行任务执行
- 通过OpenClaw通知跟踪任务进度
架构
调度(双路径)
- 1. Webhook(实时):nexum callback → POST /hooks/agent → 编排器立即唤醒
- 调度队列(备用):nexum callback → 写入nexum/dispatch-queue.jsonl → 心跳在10分钟内处理
交叉评审
- - codex-gen → claude-eval, claude-gen → codex-eval
自动路由
- - Contract YAML中的generator: auto → 系统根据任务类型选择代理
批次进度
- - nexum status显示当前批次进度及总体进度
代理命名:<模型>-<角色>-<编号>
- - codex-gen-01~03:后端/API代码
- claude-gen-01~02:面向用户的WebUI/文档
- codex-eval-01 / claude-eval-01:交叉评审
- claude-plan-01:架构(opus)
关键CLI命令
bash
nexum init [--project <目录>] [--yes]
nexum sync [任务ID] [--project <目录>]
nexum spawn <任务ID> [--project <目录>]
nexum track <任务ID> <会话密钥>
nexum callback <任务ID> [--role evaluator] [--model gpt-5.4] [--input-tokens N] [--output-tokens N]
nexum eval <任务ID>
nexum complete <任务ID>
nexum status [--project <目录>]
nexum archive [--project <目录>]
nexum health [--project <目录>]
nexum retry <任务ID> --force
Contract YAML
yaml
id: TASK-001
name: 实现功能X
batch: batch-1
agent:
generator: codex-gen-01
evaluator: claude-eval-01
scope:
files:
- src/feature.ts
deliverables:
- path: src/feature.ts
description: ...
eval_strategy:
type: review
criteria:
- id: C1
desc: ...
weight: 2
max_iterations: 3
回调协议(通过nexum init注入到AGENTS.md中)
nexum init将回调协议写入AGENTS.md作为唯一真实来源。如果项目只有CLAUDE.md,Nexum将从中生成AGENTS.md并持续更新AGENTS.md。
完成任务后,运行:
bash
nexum callback <任务ID> --project <项目目录> \
--model gpt-5.4 \
--input-tokens \
--output-tokens
Git约定
- - 直接推送到main分支,必要时可回退
- 使用英文常规提交格式:feat(scope): TASK-ID: 描述