Multi-Agent Safe-Cooperation (Single Device)

A single-device multi-Agent task chain collaboration methodology based on MGC. Provides prompt templates for Master Agent, Script Agent, and Executor Agent to achieve zero-exposure security collaboration for sensitive resources.

---

One-Line Description

A single-device multi-Agent task chain collaboration methodology based on MGC. Through Master Agent orchestration, Script Agent scripting, and Executor Agent execution, achieves zero-exposure security collaboration for sensitive resources.

---

Problem Solved

Pain Point	Solution
Multiple Agents sharing environment, keys easily leaked	Store keys in MGC, agents execute without exposure
Script content exposed to all Agents	Store scripts in MGC, zero exposure during execution
Lack of task collaboration methodology	Master Agent task chain orchestration template
Sub-Agent unauthorized access to sensitive resources	Executor Agent prompt constraints

---

Core Concepts

MGC's Role in Collaboration

MGC is the local sensitive resource hosting and execution layer:

1. User stores keys and scripts via WebUI (browser or mgc_open_webui)
2. Script Agent writes scripts and stores them in MGC via mgc_save
3. Master Agent orchestrates task chain, specifies call timing; can execute directly when sub-Agents fail or with user authorization
4. Executor Agent only calls mgc_run, never touches keys or scripts

Four Roles

Role	Responsibility	Capabilities
User	Store keys/scripts, issue commands	WebUI / mgc_open_webui
Master Agent	Task decomposition, orchestration, execute on authorization/sub-Agent failure	mgc_list, mgc_run
Script Agent	Write scripts, store in MGC	mgc_save, mgc_get, mgc_list
Executor Agent	Complete non-sensitive tasks	mgc_run, mgc_list

Note: Sensitive operations (mgc_get to retrieve plaintext) require user authorization.

---

Collaboration Flow

text

User ──Store Keys/Scripts (WebUI)──> MGC
                │
                ▼
User ──Issue Command──> Master Agent
                │
                ├──Decompose Task Chain
                │
                ▼
        ┌───────┴───────┐
        ▼               ▼
   Script Agent      Executor Agent
   (Write Scripts)  (Execute Tasks)
        │               │
        ▼               ▼
   mgc_save         mgc_run
        │               │
        └───────┬───────┘
                ▼
           MGC Execution
        (Return Results)

---

Quick Start

Prerequisites

1. Install MGC v1.4.7+
2. Configure MCP (add to AI client):

   json

   {
     "mcpServers": {
       "mgc-blackbox": {
         "command": "mgc",
         "args": ["--mcp"]
       }
     }
   }
   

3. Store sensitive resources via WebUI or mgc_open_webui:
   • API Keys
   • Database Credentials
   • Business Scripts

Step 1: Configure Master Agent

Load prompts/master_agent.md into the Master Agent's system prompt.

Step 2: Configure Script Agent

Load prompts/script_agent.md.

Step 3: Configure Executor Agent

Load prompts/executor_agent.md.

Step 4: Start Collaboration

After user issues a command, the Master Agent automatically decomposes the task chain and coordinates sub-Agents to complete.

---

Prompt Templates

prompts/master_agent.md

Master Agent prompt template, includes:

• Task decomposition methods
• MGC call timing
• Sub-task allocation rules

prompts/script_agent.md

Script Agent prompt template, includes:

• mgc_save usage
• Script naming conventions
• Script location reporting format

prompts/executor_agent.md

Executor Agent prompt template, includes:

• mgc_run usage
• Prohibited behaviors
• Result handling specifications

prompts/cooperation_best_practice.md

Best practice document, automatically maintained by Master Agent:

• Collaboration flow
• Reusable script list
• Parameter specifications
• User preferences
• Common errors and fixes
• Scenario best practices
• Collaboration records and user feedback

---

MGC Tool Usage

mgc_save (Used by Script Agent)

python

# Store script
mgc_save(
    info_type="script",
    info_owner="Data Query Script",
    ext01="python",
    content="""# Script content..."""
)

mgc_run (Used by Executor Agent)

⚠️ Note: Some MCP clients may have ext02 JSON object serialization compatibility issues; recommended to use JSON string format.

python

# Execute script
mgc_run(
    info_type="script",      # Required, specify type as script
    info_owner="Data Query Script",  # Required, script name
    diff_1="v1",             # Required, version or branch identifier
    ext02='{"param1": "value1"}'  # Optional, JSON string format for execution parameters
)

mgc_list (Available to All Agents)

Note: mgc_list only views information lists, does not retrieve plaintext; all Agents can use it.

python

# View available scripts
mgc_list(info_type="script")

mgc_open_webui (Used by User)

Used to have AI open MGC WebUI for users to store sensitive resources.

python

# Open WebUI
mgc_open_webui()

---

Security Boundaries

What This Skill Provides

• Collaboration methodology framework
• Prompt templates
• MGC tool usage specifications

What This Skill Does NOT Provide

• MGC permission enforcement (relies on prompt constraints)
• Automatic key management
• Agent identity authentication

⚠️ Important Reminders

1. All sensitive resources must be stored by users via MGC WebUI
2. Executor Agent can only call mgc_run, cannot read script content
3. Master Agent must clearly inform sub-Agents of call timing and script location
4. After sensitive operation results return, Master Agent is responsible for updating task status and coordinating sub-Agent collaboration

---

FAQ

Question	Answer
Will information stored in MGC sync to cloud?	MGC is a local tool without active network capability
What are MGC's main capabilities?	After installing MGC, view ~/.mgc/skills/mgc_skill.md
Can sub-Agents bypass MGC?	Prompt constraints cannot completely prevent; ensure sub-Agents cannot access local script files
How to isolate scripts for different tasks?	Use info_owner naming to distinguish, e.g., TaskA_QueryScript, TaskB_PublishScript
How does Master Agent know what scripts are available?	Use mgc_list to view, or report by Script Agent

---

Related Resources

• MGC Core Repository: https://github.com/zkeviny/MGC-Blackbox
• MGC Installation: pip install mgc-blackbox
• WebUI: http://127.0.0.1:57218
• API: http://127.0.0.1:57219
• MGC API Auth Token: ~/.mgc/database/mgc_black_box/.mgc_token
• Feedback: [e-mail]

---

Changelog

v1.0.0

• Initial release
• Provides Master Agent, Script Agent, Executor Agent prompt templates
• Includes complete collaboration flow documentation