caveman
claudeTalk like caveman. Cut filler. Keep technical details explicit.
22 skills · 1 command · 3 agents · hooks
Multi-agent orchestration: 9 subagents, /ulw /team /hyperplan, interview-driven planning, and an adversarial review gate. The always-on core needs no external service; graph, orchestration and dashboard are separate profiles that are off by default. Registers one UserPromptSubmit hook whose injection is gated off by default (omz.keyword_hook) — see the README before enabling it.
Plugin omz from marketplace omz-marketplace in djt889/OhMyZcode. Commands read from the repo's manifest — 1 month ago.
This plugin runs code on your machine — its hooks fire automatically on events like tool use. The exact commands are listed below; read them before installing.
review-work当一批工作完成提交前需要独立综合评审,或 omz-reviewer 单点评审不足以覆盖目标/QA/安全/上下文维度时激活。5-lane 并行评审;全 PASS 才 PASS。
ulw-execute仅当用户/主 agent 要求按已定稿的 .omz/plans/<stem>-<slug>.md 逐波执行时激活(计划缺失则先反向调 ulw-plan 补计划);不得空中执行,普通问答与单文件 quick 小改不得激活。
ulw-plan仅当用户显式说 ulw-plan、要求工作计划、或主 agent 按规划门槛(≥2 步骤/多文件/架构决策)委派规划时激活。裸的 /ulw 运行不算请求;普通'do X/fix X'在本会话粘性映射为'plan X'。未满足触发条件不得激活。
ulw-research仅当用户显式要求调研/尽调/research,或主 agent 委派饱和调研时激活。普通问答与普通查文档(Librarian 单次可查清的)不得激活本协议。
/omz:hyperplan纯规划模式:omz-planner 访谈 → omz-critic 差距分析 → 批准门等待,不执行任何实现
/omz:omz-doctorOMZ 自检:agent 可达性(spawn ping)、model 校验、gitignore、mtime(B19)、BOM 扫描(B4)
/omz:omz-statusOMZ 状态看板:渲染 .omz/ 的波次×任务×状态(40 行上限,超出聚合)
/omz:teamTeam Mode:多 worker 并行编排(coordinator MCP 或 core 文件回退)。用法 /team <目标>
/omz:ulwultrawork 模式:编排目标的全生命周期(规划→执行→双证据验证→评审门→提交)。用法 /ulw <目标>
omz-atlas当需要按已批准计划逐波驱动执行(/ulw-execute 场景)时委派。波次状态机与派单建议生成器,沟通型汇报;自己不 spawn 不实现。
omz-critic当一份计划草稿成形、尚未批准执行时委派。计划差距分析师:找遗漏场景/隐含假设/依赖风险,只评审不修改。
omz-deep当任务属于 deep 类(棘手调试/研究密集/微妙跨模块)且需要端到端自主实现时委派。给目标不给步骤的深度自主编码者;范围清晰的标准任务勿派,用 junior。
omz-junior当任务可独立完成、范围清晰、属单 lane 执行类(标准特性/文档/UI 组件等)时委派。聚焦单任务执行器,禁止再委派;单文件 typo 级小改勿派,主 agent 自己干。
omz-librarian当需要查外部文档/API 用法/版本兼容/第三方库资料时委派。检索员:按 URL 抓全文,输出带来源引用的结论。
omz-looker当需要视觉分析(截图/图表/渲染工件)或 visual-QA 门时委派。多模态检查员:只看图,输出逐图判定;须给全图片路径清单。
omz-oracle当遇到架构决策/疑难调试/需要第二意见的技术判断时委派。资深架构顾问:只分析给方案,不动任何代码。
omz-planner当任务满足规划门槛(≥2 步骤/多文件/含架构决策)或用户要求工作计划/访谈时委派。访谈式规划顾问:产出分波次计划,自己绝不实现;单步骤单文件任务勿派。
omz-reviewer当评审门触发(措辞严格/≥3 文件/≥20 轮/≥30 分钟/重构迁移性能安全)或需 AdversarialVerify 裁决 DoneClaim 时委派。独立只读对抗评审,分级 blocker/major/minor;门未触发勿派。
node
Talk like caveman. Cut filler. Keep technical details explicit.
22 skills · 1 command · 3 agents · hooks
Persistent memory for Claude Code: the next session starts where the last one left off.
29 skills · 1 MCP server · hooks
tasteskill 2.5.1: professional company websites at the figma.com and x.ai level, with one section per screen, real photographs, real product UI, effects kept for tech and gaming pages, and a real-browser checker and finish gate.
1 skill
Open-source AI job search that runs locally in your AI coding CLI: scan job portals, evaluate listings into a structured A-H report with a 1-5 score, tailor your CV, track applications
5 skills · hooks
Cross-session memory and token savings for coding agents.
6 skills · 1 agent · 1 MCP server · hooks
Package any website or local web build into a lightweight desktop app with Pake CLI.
1 skill
This page updates itself from your repo. Add the badge to your README, or get it Featured to pin it above the organic results with a followed link to your site.