Team Builder

在 OpenClaw 上一键部署多 Agent SaaS 团队工作区。内置双开发轨（devops 交付 + fullstack-dev 实现）、实时 spawn 调度、cron 巡检、Deep Dive 产品知识目录、onboarding 引导。支持自定义角色、模型、时区，可选 Telegram 接入。

Before installing or running this skill: - Review scripts/deploy.js and the generated apply-config.js and create-crons.* files (the README & SKILL.md tell you where they will be written). Do not run apply-config.js or create-crons scripts until you have inspected them. - Back up your existing ~/.openclaw/openclaw.json before using this skill; the generator reads that file and the generated apply-config.js will write to it. - If you plan to provide Telegram bot tokens, understand they will be stored in openclaw.json (a global config file). Consider whether you want those tokens in that file and who can read it. - The skill may read project directories for Deep Dive scans (git, grep, file reads). Limit where you run these scans and avoid running them on sensitive repos without inspection. - Use the --verify mode and test in a throwaway workspace (small team, separate workspaceDir) first to observe what files are created. - If you need more assurance, request the author/source or a signed release; the package has no homepage/source provenance listed, which lowers supply-chain confidence.

Type: OpenClaw Skill Name: team-builder Version: 3.0.0 The 'team-builder' skill is a deployment framework that automates the setup of a multi-agent workspace, but it employs several high-risk behaviors. The core script `scripts/deploy.js` generates an `apply-config.js` file designed to modify the global `openclaw.json` configuration, adding eight new agents and enabling broad `agentToAgent` communication permissions. It also creates ten persistent cron jobs via `create-crons.sh/ps1` for automated agent execution. Furthermore, the 'Deep Dive' protocol (found in `references/agent-refs/fullstack-dev/deep-dive-protocol.md`) instructs agents to perform extensive scans of local project directories, utilizing system commands like `tree`, `grep`, and framework-specific CLIs to extract architectural and database metadata. While these capabilities are consistent with the stated goal of building an autonomous development team, the combination of global configuration modification, persistence, and broad filesystem discovery represents a significant security surface.