Skill Evolution Driver

Task Objectives

This Skill is used to: Drive self-evolution of skills, continuously optimize as the project progresses
Capabilities include:
1. Monitor data stored in skill-manager to identify skill optimization opportunities
2. Notify users of potential optimization suggestions during idle time
3. Maintain a list of skill optimization tasks
4. Execute skill update processes (backup, update, testing, restoration)
5. Manage skill version numbers
Trigger conditions: Triggered by any of the following situations
- Direct user requests: "Optimize skill", "Improve skill", "Skill needs update", "Skill upgrade", "Fix skill issues"
- Automatic monitoring trigger: Detects pending optimization tasks in skill data stored in skill-manager (such as optimization tasks with status=pending, errors and warnings in logs, missing required fields)
- Version-related: "Skill version number needs update", "Need to record skill changes"
- Quality-related: "Skill format is incorrect", "Need to check skill quality", "Skill test failed"

Optimization Identification

Data Sources

Analyze optimization opportunities from skill data stored in skill-manager:

Missing information: Lack of required fields in skill configuration (such as version)
Format issues: Data format does not comply with specifications
Duplicate information: Redundant or duplicate content exists
Usage patterns: Identify improvement points based on access frequency and methods
Error logs: Identify common errors or warnings from logs
Cross-skill collaboration issues: Identify dependency issues or improvement suggestions for the current skill from logs of other skills

Optimization Types

format_improvement: Format improvement
content_optimization: Content optimization
version_update: Version number update
bug_fix: Bug fix
feature_enhancement: Feature enhancement

Evolution Process

Step 1: Notify User

Automatic Monitoring Trigger

The agent periodically checks data stored in skill-manager to identify optimization opportunities:

Inspection Tools: Call

scripts/check_optimization_opportunities.py

bash

# Check all skills
python scripts/check_optimization_opportunities.py

# Check specific skill
python scripts/check_optimization_opportunities.py --skill-name dev-observability

Inspection Content:

Pending tasks: Check optimization tasks stored in skill-evolution-driver itself (status=pending)
Missing fields: Check if skill configuration lacks required fields (such as version, deploy_mode, manual_path)
Error logs: Check ERROR/WARNING/CRITICAL logs in skill logs
Directory issues: Check if skill directory structure complies with specifications

Notification Format:

Potential optimization opportunities detected:
- Skill: skill-name
- Optimization type: format_improvement
- Description: SKILL.md is missing version field
- Start optimization? (y/n)

User-Initiated Trigger

Directly notify the user when the user expresses the following needs:

"Optimize skill", "Improve skill"
"Skill needs update", "Skill upgrade"
"Fix skill issues"
"Skill version number needs update"
"Skill format is incorrect", "Need to check skill quality"

Step 2: Maintain Task List

If the user agrees, call skill-manager to store the optimization task:

json

{
  "task_id": "OPT-001",
  "skill_name": "skill-name",
  "optimization_type": "format_improvement",
  "description": "SKILL.md is missing version field",
  "status": "pending",
  "feasibility": "pending",
  "backup_path": "",
  "old_version": "",
  "new_version": "",
  "test_result": "",
  "notes": "",
  "created_at": "2024-01-22 12:00:00",
  "updated_at": "2024-01-22 12:00:00"
}

Storage method: Update the configuration of skill-evolution-driver in skill-manager

python

import sys
sys.path.insert(0, '/workspace/projects/skill-manager/scripts')
from skill_manager import SkillStorage

storage = SkillStorage(data_path="/workspace/projects/skill-data.json")
existing_config = storage.get_config("skill-evolution-driver") or {}
existing_config["optimization_tasks"] = [task_dict]  # Task list
storage.save_config("skill-evolution-driver", existing_config)

Step 3: Analyze Feasibility

Analyze the feasibility of optimizing the skill:

Feasibility Evaluation Criteria:

Complete skill directory structure
Correct SKILL.md format
Valid script syntax
Clear dependency relationships
Clear optimization content

Infeasible Scenarios:

Skill directory does not exist or is damaged
SKILL.md format is incorrect and cannot be parsed
Optimization content is unclear or overly complex
Optimization may damage existing functions

Update task status:

Feasible:
```
feasibility: "feasible"
```
Infeasible:
```
feasibility: "not_feasible"
```
, and explain the reason in
```
notes
```

Step 4: Backup Skill

Call

scripts/backup_skill.py

to backup the skill:

bash

python scripts/backup_skill.py --skill-dir /workspace/projects/skill-name

Backup file format:

skill-name.backup.<timestamp>.skill

Update task:

backup_path: "/workspace/projects/skill-name.backup.<timestamp>.skill"

Step 5: Update Skill

Execute specific optimization operations and update the version number:

Execute optimization: Modify skill content according to optimization type
Update version number: Call
```
scripts/update_version.py
```

bash

python scripts/update_version.py --skill-dir /workspace/projects/skill-name --type patch

Version number types:

```
patch
```
: Patch version (v1.0.0 -> v1.0.1)
```
minor
```
: Minor version (v1.0.0 -> v1.1.0)
```
major
```
: Major version (v1.0.0 -> v2.0.0)

Update task:

```
old_version: "v1.0.0"
```
```
new_version: "v1.0.1"
```

Step 6: Conduct Testing

Test and verify the updated skill:

Testing Content:

SKILL.md format verification (YAML preamble area)
Script syntax check (Python syntax)
Directory structure verification (complies with fixed structure)
Dependency integrity check (dependency field)

Test script:

bash

python scripts/backup_skill.py --validate-only --skill-dir /workspace/projects/skill-name

Step 7: Handle Test Results

Test Failed:

Call
```
scripts/restore_skill.py
```
to restore the skill
Use
```
manage_optimization_tasks.py
```
to update task status

bash

# Restore skill
python scripts/restore_skill.py --backup-file <backup path> --skill-dir /workspace/projects/skill-name

# Update task status
python scripts/manage_optimization_tasks.py update \
  --task-id OPT-001 \
  --status failed \
  --test-result failed \
  --notes "Test failed: SKILL.md format error"

Test Passed:

Use
```
manage_optimization_tasks.py
```
to update task status
Package the new .skill file
Inform the user of update content
Remind the user to reload the skill

bash

# Update task status
python scripts/manage_optimization_tasks.py update \
  --task-id OPT-001 \
  --status completed \
  --test-result passed \
  --notes "Optimization successful: version field added" \
  --new-version v1.0.1 \
  --backup-path /workspace/projects/skill-name.backup.20260122120000.skill

Important: Must update the task status to

completed

to avoid repeated reminders of the same optimization opportunity during next inspection.

Step 8: Notify User

Notify the user based on test results:

Test Passed:

✓ Skill optimization successful
- Skill: skill-name
- Version: v1.0.0 -> v1.0.1
- Update content: Added version field
- Backup: skill-name.backup.<timestamp>.skill

Please reload the AI interaction interface to use the updated skill.

Test Failed:

✗ Skill optimization failed
- Skill: skill-name
- Reason: Test failed
- Details: SKILL.md format error
- Skill has been restored

Core Function Description

Functions Handled by the Agent

Monitor skill-manager data to identify optimization opportunities
Notify users of optimization suggestions during idle time
Maintain skill optimization task list
Analyze optimization feasibility
Execute optimization operations
Conduct test verification
Handle test results
Notify users

Functions Implemented by Scripts

Check optimization opportunities:
```
scripts/check_optimization_opportunities.py
```
scans skill-manager data to automatically identify optimization opportunities (filter completed tasks)
Manage optimization tasks:
```
scripts/manage_optimization_tasks.py
```
provides add/delete/update/query functions for optimization tasks
Backup skill:
```
scripts/backup_skill.py
```
backs up the entire skill directory
Restore skill:
```
scripts/restore_skill.py
```
restores skill from backup
Update version number:
```
scripts/update_version.py
```
updates the version field in SKILL.md
Test verification:
```
scripts/backup_skill.py --validate-only
```
verifies skill format

Resource Index

Optimization opportunity inspection tool: See scripts/check_optimization_opportunities.py (automatically scans skill-manager data to identify optimization opportunities and filter completed tasks)
Optimization task management tool: See scripts/manage_optimization_tasks.py (add/delete/update/query of optimization tasks)
Backup tool: See scripts/backup_skill.py (backs up skill directory)
Restore tool: See scripts/restore_skill.py (restores skill from backup)
Version management: See scripts/update_version.py (updates version number)
Optimization guide: See references/optimization_guide.md (optimization types, test standards, best practices)

Notes

Must backup the skill before optimization
Keep task status synchronized during optimization process
Must restore the skill if test fails
Version numbers follow semantic versioning specifications
Major changes require explicit user confirmation
Must update task status to
```
completed
```
after optimization to avoid repeated reminders
Remind user to reload the skill after optimization

Best Practices

Regularly monitor skill-manager data to identify optimization opportunities
Prioritize high-value, low-risk optimizations
Fully analyze feasibility before optimization
Maintain detailed optimization logs
Regularly clean up expired backup files
Provide clear optimization instructions to users

Usage Examples

Example 1: Add version field

Detected that SKILL.md of skill-name is missing version field.

Notify user
Create optimization task after user agrees
Analyze feasibility: feasible
Backup skill: skill-name.backup.20240122120000.skill
Update SKILL.md, add version: v1.0.0
Test verification: passed
Update task status: completed
Notify user: optimization successful

Example 2: Format improvement

Detected that SKILL.md of skill-name does not comply with format specifications.

Notify user
Create optimization task after user agrees
Analyze feasibility: feasible
Backup skill
Correct SKILL.md format
Test verification: passed
Update task status: completed
Notify user: optimization successful

Example 3: Infeasible optimization

Detected that optimization content of skill-name is overly complex.

Notify user
Create optimization task after user agrees
Analyze feasibility: infeasible
Update task status: failed
Update notes: optimization content is overly complex, requires manual intervention
Notify user: optimization is infeasible, manual processing required

skill-evolution-driver

NPX Install

Tags

SKILL.md Content (Chinese)

Skill Evolution Driver

Task Objectives

Optimization Identification

Data Sources

Optimization Types

Evolution Process

Step 1: Notify User

Automatic Monitoring Trigger

User-Initiated Trigger

Step 2: Maintain Task List

Step 3: Analyze Feasibility

Step 4: Backup Skill

Step 5: Update Skill

Step 6: Conduct Testing

Step 7: Handle Test Results

Step 8: Notify User

Core Function Description

Functions Handled by the Agent

Functions Implemented by Scripts

Resource Index

Notes

Best Practices

Usage Examples

Example 1: Add version field

Example 2: Format improvement

Example 3: Infeasible optimization