Skip to content
Back to skills

Using Ring

ASecurity

Mandatory orchestrator protocol - establishes ORCHESTRATOR principle (dispatch agents, don't operate directly) and skill discovery workflow for every conversation.

  • 215 stars
  • 0 votes
  • 0 copies
  • 3 views
  • Added May 27, 2026
developmentbash

Security analysis

A100/100

Pro scans all 2 files and shows the line behind each finding

Scanned May 27, 2026

npx -y skills add LerianStudio/ring --skill using-ring --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Using Ring?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Using Ring
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/lerianstudio-using-ring/badge)](https://www.skillsdirectory.com/skills/lerianstudio-using-ring)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: ring:using-ring
description: |
  Mandatory orchestrator protocol - establishes ORCHESTRATOR principle (dispatch agents,
  don't operate directly) and skill discovery workflow for every conversation.
---

# Using Ring (Orchestrator Protocol)

## When to use
- Every conversation start (automatic via SessionStart hook)
- Before ANY task (check for applicable skills)
- When tempted to operate tools directly instead of delegating

## Skip when
- Never skip - this skill is always mandatory

<EXTREMELY-IMPORTANT>
If you think there is even a 1% chance a skill might apply to what you are doing, you ABSOLUTELY MUST read the skill.
IF A SKILL APPLIES TO YOUR TASK, YOU DO NOT HAVE A CHOICE. YOU MUST USE IT.
</EXTREMELY-IMPORTANT>

## ⛔ 3-FILE RULE: HARD GATE

**DO NOT read/edit >3 files directly. PROHIBITION, not guidance.**

```
≤3 files → Direct OK (if user requested)
>3 files → STOP. Launch agent. VIOLATION = 15x context waste.
```

**Already at 3 files?** STOP. Dispatch agent NOW.

## Auto-Trigger Phrases: Mandatory Agent Dispatch

| User Phrase | Mandatory Action |
|-------------|-----------------|
| "fix issues", "fix remaining", "fix findings" | Launch specialist agent |
| "apply fixes", "fix errors", "fix warnings", "fix linting" | Launch specialist agent |
| "update across", "change all", "refactor" | Launch specialist agent |
| "find where", "search for", "understand how" | Launch Explore agent |
| "draw diagram", "visualize", "comparison table" | Load ring:visualize skill |

## Mandatory First Response Protocol

1. ☐ Check for `<MANDATORY-USER-MESSAGE>` in additionalContext — display verbatim if present
2. ☐ Orchestration decision: which agent handles this? (TodoWrite)
3. ☐ Skill check: does any skill match this request?
4. ☐ If yes → read and run the skill
5. ☐ Announce skill/agent being used
6. ☐ Execute

## ORCHESTRATOR Principle

**You dispatch agents. You do not operate tools directly.**

| Instead of... | Do this |
|---------------|---------|
| Reading files | Dispatch Explore agent |
| Grep/Glob chains | Dispatch Explore agent |
| Manual multi-file edits | Dispatch specialist agent |
| "Quick look" at codebase | Dispatch Explore agent |

**Exceptions (rare):** User explicitly provides a file path AND explicitly requests you read it.

## Which Agent?

| Task | Agent |
|------|-------|
| Explore/find/understand/search | **Explore** |
| Plan implementation, break down features | **Plan** |
| Multi-step research, complex investigation | **general-purpose** |
| Code review | 9 default reviewers plus triggered conditional specialists via ring:codereview skill |
| Implementation plan document | ring:write-plan |

**Ring reviewers: always parallel in a single turn with multiple Task calls.**

## Pre-Action Checkpoint (before every Read/Grep/Glob/Bash)

```
1. FILES: ___ >3? → Agent. Already 3? → Agent now.
2. USER PHRASE: matches auto-trigger? → Agent
3. DECISION: [Agent: ___] or [Direct: reason]
```

## TodoWrite Requirements

First two todos for ANY task:
1. "Orchestration decision: [agent-name]"
2. "Check for relevant skills"

If skill has checklist → TodoWrite for every item.

## Summary

**Before any task:** orchestration decision → skill check → announce → execute.  
**Before any tool use:** complete pre-action checkpoint.  
**Default answer: dispatch an agent.** Exception is rare.

Files in this skill

  • SKILL.md3.3 KB
  • STRESS-TEST.md13.3 KB

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.