AI Manager · Five Guardians
Paste the following prompt into your AI chat to install this skill:
Install @user_e23813cc/ai-guandashi-enterprise-v2 by following https://skillhub.cn/install/skillhub.md.
About this skill
Problem
Small teams often split work across multiple agents and then hit role overreach, lost tasks, channel confusion, and uncontrolled cost or concurrency. AI Manager · Five Guardians tries to place Grand Steward, Housekeeper, Growth, Procurement, Talent, Finance, and Super Assistant into one dispatch system instead of letting each agent take work independently.
How It Works and Limits
It separates responsibilities: Grand Steward handles assignment and progress tracking, while business agents execute customer acquisition, procurement, HR, finance, and admin actions through todo folders and sessions_send. Core workflows cover deal closing, procurement and payment, onboarding and offboarding, monthly operations review, and risk alerts. Check hard limits before use:
- One agent supports up to
5concurrent tasks, the system up to20 - Single tasks auto-fuse after
30minutes, and tool calls have daily cost caps - Files must follow UTF-8, path, and format rules, and financial operations require secondary confirmation
- Data stays in the local
workspaceby default, and sensitive data must not move between unnecessary agents
Use Cases
- A boss sends a deal task to Grand Steward on WeChat and expects agents to connect growth and housekeeping steps.
- Procurement submits a supplier payment request, finance drafts budget and risk notes, then the steward asks for confirmation.
- During offboarding, HR builds a handover list, finance checks payroll and expenses, and the steward archives the todo file.
- An operations lead drops a UTF-8 Markdown customer list into the todo folder for steward dispatch to business agents.
Best For
- SMB owners who need WeChat, WeCom, and file tasks under one dispatch system.
- Ops leads who want to split growth, procurement, finance, and HR across agents with daily cost caps.
- Automation engineers building local QClaw/OpenClaw agent workflows and keeping data in the workspace.
- Managers who need secondary confirmation and audit trails for high-risk workflows like offboarding and payments.
Related Skills
Automatically searches job postings based on the user profile, AI-scores fit, saves desktop reports, and sends email updates with scheduled tracking.
Activates bionic reasoning for causal judgment, numerical prediction, and hypothesis validation, using hypothesis-driven checks, Bayesian updates, falsifiability tests, bias defense, and physical constraints.
Triggered by /plan, it asks the agent to output a plan, risks, impact scope, and validation approach before execution.
Track and clean agent session files, packages, and Skills via trash-first safety.