🧠 Intelligent Memory Trigger System#

🎯 Skill Overview#

An intelligent system that automatically determines when to create workflow documentation based on task complexity, repetition patterns, and user intent. Transforms agents from "passively waiting for instructions" to "actively providing help."

📋 Usage Scenarios#

Automatically triggers this skill when encountering the following situations:

Mandatory Trigger Conditions (Hard Requirements)#

1. Steps ≥ 3: Task contains 3 or more operational steps
2. Cross-system ≥ 2: Task involves 2 or more system components
3. Configuration Changes: Modifying core configuration files or system settings
4. Repetition Occurrence: Same or similar task appears for the 2nd time
5. User Explicit Request: User indicates "remember", "summarize", "process"

Recommended Trigger Conditions#

1. Complexity Score ≥ 6 points (see evaluation algorithm)
2. Potential Reuse: User mentions "next time", "in the future"
3. Team Collaboration Tasks: Involving multiple agent collaboration
4. Error Risk: Complex operations prone to errors

🧠 Core Decision Logic#

Complexity Evaluation Algorithm#

javascript

function evaluateComplexity(taskDescription) {
  let score = 0;

  // 1. Step Count Evaluation
  if (taskDescription contains "first step" and "second step") score += 2;
  if (taskDescription contains "third step") score += 2;
  if (taskDescription contains "fourth step" or "finally") score += 2;

  // 2. System Involvement Evaluation
  if (taskDescription contains "configuration file" or ".json") score += 1;
  if (taskDescription contains "command line" or "command") score += 1;
  if (taskDescription contains "restart" or "gateway") score += 1;
  if (taskDescription contains "Feishu" or "WeChat" or "platform") score += 1;

  // 3. Configuration Impact Evaluation
  if (taskDescription contains "configuration" or "settings") score += 1;
  if (taskDescription contains "modify" or "edit") score += 1;
  if (taskDescription contains "add" or "delete") score += 1;

  // 4. Collaboration Requirement Evaluation
  if (taskDescription contains "team" or "collaboration") score += 1;
  if (taskDescription contains "multiple" or "several") score += 1;

  return score;
}

Decision Thresholds#

• Total Score ≥ 8 points → Must create memory trigger
• Total Score 6-7 points → Recommend creating memory trigger
• Total Score ≤ 5 points → No need to create memory trigger

🔍 Repetition Pattern Detection#

Task Type Identification Keywords#

text

System Configuration: ["configuration", "settings", "install", "deploy"]
File Operations: ["create", "edit", "delete", "move"]
Command Line: ["execute", "run", "command", "script"]
Platform Integration: ["Feishu", "WeChat", "API", "integration"]

Repetition Detection Logic#

When similarity between current task and historical tasks > 0.7, treat as repetition pattern and should create or reference existing workflow.

🎯 User Intent Analysis#

Explicit Request Keywords#

text

Recording Requests: ["remember", "record", "save", "memo"]
Summary Requests: ["summarize", "organize", "arrange", "sort"]
Process Requests: ["process", "steps", "method", "operation"]
Future Use: ["next time", "in the future", "later", "repeat"]

When detecting these keywords, proactively ask if workflow documentation is needed.

💬 Proactive Inquiry Templates#

Based on Complexity#

"Teacher Sun, this task appears to be relatively complex (score: X points) and may be reused. Would you like me to create a workflow document?"

Based on Repetition Pattern#

"Teacher Sun, this task is very similar to the previous [similar task]. Should I reference the previous workflow?"

Based on User Intent#

"Teacher Sun, you mentioned needing to [record/summarize/process]. Should I create the document now?"

Based on Configuration Importance#

"Teacher Sun, this configuration change may affect system stability. Should I record the detailed steps?"

📝 Workflow Document Creation#

Document Naming Convention#

• [task-type]-workflow.md
• Example: feishu-bot-configuration-workflow.md

Document Structure Requirements#

1. Workflow Overview: Task objectives and value
2. Configuration Steps: Detailed operational steps (clearly numbered)
3. Verification Methods: How to verify successful configuration
4. Common Issues: Potential problems and solutions
5. Related Files: Involved configuration files and documents

Storage Location#

• Main Directory: C:\Users\sjh65\.openclaw\workspace\workflows\
• Quick Access: Dedicated documents in root directory

🔧 System Maintenance#

Regular Checks#

• Weekly: Check task history, identify high-complexity tasks without workflows
• Monthly: Optimize complexity scoring standards
• Quarterly: Review and update all workflow documents

Optimization Mechanism#

• Adjust trigger conditions based on user feedback
• Optimize keywords based on actual usage
• Improve workflow documents based on task execution effectiveness

📈 Effectiveness Evaluation Metrics#

Quantitative Metrics#

1. Workflow Coverage: Proportion of high-complexity tasks with established workflows
2. Proactive Inquiry Rate: Proportion of proactively asking to create workflows when needed
3. User Reminder Rate: Proportion needing user reminders to provide workflows (goal: reduce)
4. Workflow Usage Rate: Frequency of workflow documents being actually referenced

Quality Metrics#

1. Workflow Completeness: Whether steps are complete and clear
2. Trigger Accuracy: Whether keyword triggers are accurate
3. User Satisfaction: Whether workflows truly help with work

🎯 Practical Application Cases#

Case 1: Feishu Bot Configuration#

• Complexity Score: 9 points (steps 4 + cross-system 3 + system impact 2)
• Trigger Conditions: Cross-system ≥ 2, configuration changes, repetition occurrence
• Result: Created feishu-bot-configuration-workflow.md
• Trigger Keywords: ["Feishu bot", "configure bot", "channel configuration", "pairing"]

Case 2: File Reading#

• Complexity Score: 3 points
• Trigger Conditions: None
• Result: No need to create workflow document

Case 3: Model Testing#

• Complexity Score: 6 points
• Trigger Conditions: Complexity score ≥ 6 points
• Result: Asked "Would you like me to create a workflow document?"

🔄 Complete Workflow#

text

Start Task
    ↓
Analyze Task Description
    ↓
Evaluate Complexity (step count, system involvement, configuration impact)
    ↓
Check Task History (is it a repetitive type?)
    ↓
Analyze User Intent (are there "remember", "summarize" keywords?)
    ↓
    ┌─────────────────────┐
    │ Need to create memory trigger? │
    └─────────────────────┘
            ↓
    ┌─────────────────────┐
    │ Yes → Proactively ask user │
    │ No → Execute task normally │
    └─────────────────────┘

📋 Quick Reference#

Must Create Workflow Situations#

1. Steps ≥ 3
2. Cross-system ≥ 2
3. Configuration changes
4. Task repeats for 2nd time
5. User explicit request
6. Complexity score ≥ 8 points

Recommend Inquiry Situations#

1. Complexity score 6-7 points
2. Potential reuse
3. Team collaboration tasks
4. Error risk

No Need for Workflow Situations#

1. Complexity score ≤ 5 points
2. Clearly one-time operations
3. Simple queries or information retrieval
4. Exploratory trial-and-error operations

---

Skill Creation Time: 2026-03-18 Upgraded to English Version: 2026-03-20 Original Creator: Commander (based on Shrimp Assistant's design) Upgraded By: Shrimp Assistant (main agent) Goal: Achieve systematic, automated memory trigger decision-making