Назад към всички

skill-priority-setup

// Scans installed skills, suggests L0-L3 priority tiers, and auto-configures skill injection policy. Use when: setting up skill priorities, optimizing token budget, or migrating to tiered skill architecture.

$ git log --oneline --stat
stars:1,933
forks:367
updated:March 4, 2026
SKILL.mdreadonly
SKILL.md Frontmatter
nameskill-priority-setup
descriptionScans installed skills, suggests L0-L3 priority tiers, and auto-configures skill injection policy. Use when: setting up skill priorities, optimizing token budget, or migrating to tiered skill architecture.
version1.0.0
authordeepeye
tagsskills,priority,setup,configuration,token-optimization
clawhubhttps://clawhub.ai/halfmoon82/skill-priority-setup

Skill Priority Setup

Automated skill tiering and injection policy configuration for OpenClaw.

What This Skill Does

  1. Discovery Phase: Scans all installed skills across standard directories
  2. Analysis Phase: Suggests L0-L3 tier assignments based on skill characteristics
  3. Configuration Phase: Applies the tiered architecture to your setup
  4. Validation Phase: Verifies configuration and estimates token savings

Quick Start

# Run the interactive setup wizard
python3 ~/.openclaw/workspace/skills/skill-priority-setup/scripts/setup.py

# Or non-interactive mode with defaults
python3 ~/.openclaw/workspace/skills/skill-priority-setup/scripts/setup.py --auto

When to Use

  • New Setup: Just installed multiple skills and want optimal configuration
  • Performance Issues: High token usage or slow responses
  • Migration: Upgrading from flat skill structure to tiered architecture
  • Audit: Reviewing and optimizing existing skill priorities

Architecture Overview

This skill implements a 4-tier priority system:

L0 - ROM Core (Always Active)

  • semantic-system: Semantic routing (message injector)
  • agent-evolution: Self-improvement behaviors (SOUL.md)
  • config-modification: Config safety (on-demand)
  • skill-safe-install: Installation safety (on-demand)

L1 - Routing Layer (Task-Triggered)

  • browser-automation: Web automation
  • find-skills: Skill discovery
  • teamtask: Multi-agent coordination

L2 - Domain Skills (Keyword-Triggered)

  • Document: word-docx, tesseract-ocr
  • Media: youtube-transcript
  • Platform: discord, wechat-suite, evomap
  • Automation: automation-workflows

L3 - Extensions (Manual/On-Demand)

  • Third-party integrations: notion, slack, github, etc.

Workflow

┌─────────────────┐
│ 1. Scan Skills  │ → Find all SKILL.md files
└────────┬────────┘
         ▼
┌─────────────────┐
│ 2. Analyze      │ → Detect skill types & dependencies
└────────┬────────┘
         ▼
┌─────────────────┐
│ 3. Suggest Tiers│ → Propose L0-L3 assignments
└────────┬────────┘
         ▼
┌─────────────────┐
│ 4. User Review  │ → Confirm or modify suggestions
└────────┬────────┘
         ▼
┌─────────────────┐
│ 5. Apply Config │ → Update AGENTS.md, SOUL.md, etc.
└────────┬────────┘
         ▼
┌─────────────────┐
│ 6. Validate     │ → Check JSON, test injection
└─────────────────┘

Token Budget

  • L0 Core: ≤300 tokens/round (minimal rules)
  • L1 Triggered: ≤400 tokens per injection
  • Total Budget: ≤900 tokens/round
  • Overflow Strategy: Inject summary + file path only

Safety Features

  • Backup before any configuration changes
  • JSON validation before applying
  • Dry-run mode (--dry-run)
  • Rollback capability

Files Created

  • SKILL_PRIORITY_POLICY.md: Your custom policy document
  • AGENTS.md updates: Core constraints added
  • SOUL.md updates: Agent evolution behaviors
  • Backup files: .backup/*.timestamp

Configuration

Edit ~/.openclaw/workspace/skills/skill-priority-setup/config.yaml to customize:

# Default tier assignments (override per skill)
default_tiers:
  semantic-router: L0
  browser: L1
  word-docx: L2
  
# Token budgets
token_budget:
  l0_max: 300
  l1_max: 400
  total_max: 900
  
# Safety settings
backup_before_change: true
validate_json: true
auto_rollback: true

Troubleshooting

Gateway Won't Start

  • Check if message injector content exceeds limit
  • Run with --diagnose flag

Skills Not Loading

  • Verify allowBundled list in openclaw.json
  • Check file permissions

High Token Usage

  • Review L0 assignments (should be minimal)
  • Consider moving skills to lower tiers

References