← 返回 Skills 市场
OpenClaw Projects
作者
EncryptShawn
· GitHub ↗
· v1.0.0
· MIT-0
44
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install openclaw-projects
功能描述
Builds a Project — a coordinated multi-agent team setup — inside OpenClaw, for any kind of team: software development, marketing, real estate, content, sales...
安全使用建议
This skill appears to be what it says: it will interview you, generate a full project folder in ~/.openclaw/projects/, write many coordination files, update participating agents' AGENTS.md blocks, and update OpenClaw configuration to enable agent-to-agent communication. Before installing or running it: 1) Verify you trust the openclaw-administrator skill this one requires — that dependency has the privilege to change OpenClaw config and is the primary escalation risk. 2) Back up your existing ~/.openclaw/ config and AGENTS.md files so you can revert changes. 3) Confirm that you want the skill to perform file writes and config edits autonomously; if not, require manual review/approval during the interview flow. 4) Ensure your Asana/ClickUp integration and credentials remain managed by the dedicated dependency skills (this skill claims it will not hold credentials). 5) Consider running the skill in a test environment or with non-critical agents first to observe its produced files and config changes.
功能分析
Type: OpenClaw Skill
Name: openclaw-projects
Version: 1.0.0
The skill automates the setup of multi-agent project environments, requiring high-privilege actions such as modifying agent workspaces and updating global 'agent_to_agent' communication allow-lists via the openclaw-administrator skill (SKILL.md, Step 9). While these capabilities are directly aligned with the stated purpose of enabling agent coordination and include safety guardrails like mandatory user approval and recovery snapshots, the ability to alter security configurations and cross-agent access permissions constitutes a broad and risky set of permissions. No evidence of malicious intent, data exfiltration, or obfuscation was found in the provided files (SKILL.md, references/templates.md).
能力标签
能力评估
Purpose & Capability
The skill's name and description match what it does: it runs a structured interview, generates project files under ~/.openclaw/projects/[project-id], updates participating agents' AGENTS.md blocks, and updates OpenClaw config to wire agent-to-agent communication. Requiring the openclaw-administrator skill to modify config is consistent with the stated functionality.
Instruction Scope
SKILL.md explicitly instructs the agent to create many files (PROJECT.md, project.json, queues, project-lock.json, STATE.md, etc.) and to edit each participating agent's AGENTS.md and OpenClaw config. Those file- and config-modification actions are expected for this skill but are sensitive operations because they affect agent behavior and coordination. The skill claims it will not hold task-manager credentials or call Asana/ClickUp APIs directly — instead it delegates to other installed dependency skills, which is consistent with the instructions.
Install Mechanism
This is an instruction-only skill with no install spec and no bundled code — the lowest-risk install mechanism. Nothing is downloaded or written by an installer beyond what the skill instructs at runtime.
Credentials
The skill declares no required environment variables, binaries, or credentials. That aligns with its claim that task-manager API calls and any credential storage are handled by separate dependency skills. The one relevant external requirement is that the openclaw-administrator skill be installed/loaded so this skill can update config; this is proportionate but introduces a trust dependency on that administrator skill.
Persistence & Privilege
The skill does not set always:true. It can be invoked autonomously (disable-model-invocation is false), and its runtime instructions perform persistent changes to the user's home config and per-agent workspaces. This is coherent for a project-creation tool, but because it modifies config and agent state, you should only enable it when you trust the openclaw-administrator dependency and are comfortable with the skill making those persistent changes autonomously.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install openclaw-projects - 安装完成后，直接呼叫该 Skill 的名称或使用
/openclaw-projects触发 - 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v1.0.0
Initial release: Adds coordinated multi-agent project setup to OpenClaw.
Support for projects is what OpenClaw has been missing, now you can build a cohesive multi agent team to do anything, as a team!
- Introduces the Project concept, enabling teams of agents to work together with defined workflows, team rulebooks, configuration, and inter-agent messaging.
- Supports any team type (software, marketing, real estate, etc.) and integrates with Asana or ClickUp via separately-installed dependency skills.
- Guides users through a structured interview (team identity, work structure) and auto-generates a comprehensive project plan before creation.
- Builds the entire project folder structure and updates OpenClaw config for agent collaboration.
- Ensures credential security by never handling credentials; delegates all task manager API operations to dependency skills.
- Requires openclaw-administrator; recommends openclaw-recovery-manager for safety and rollback. Also Asana tool or Clickup skills (independent of this skill) for project coordination
元数据
常见问题
OpenClaw Projects 是什么？
Builds a Project — a coordinated multi-agent team setup — inside OpenClaw, for any kind of team: software development, marketing, real estate, content, sales... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 44 次。
如何安装 OpenClaw Projects？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install openclaw-projects」即可一键安装，无需额外配置。
OpenClaw Projects 是免费的吗？
是的，OpenClaw Projects 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
OpenClaw Projects 支持哪些平台？
OpenClaw Projects 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。
谁开发了 OpenClaw Projects？
由 EncryptShawn（@encryptshawn）开发并维护，当前版本 v1.0.0。
推荐 Skills