Add New Skill to lc-essentials Plugin

You are creating a new skill for the lc-essentials Claude Code plugin based on the user's description below.

User's Skill Description

The user wants a skill that: $ARGUMENTS

Your job is to create a skill that fulfills this description while conforming to the established framework patterns and conventions.

Step 1: Research Claude Code Skills Best Practices

First, use the Task tool with subagent_type="claude-code-guide" to look up:

• Official Claude Code documentation on creating skills (SKILL.md format)
• Best practices for skill descriptions and discovery optimization
• How skills interact with sub-agents and the Task tool

Step 2: Review Plugin Framework Documentation

Read these files to understand the lc-essentials framework (paths relative to plugin root):

• CALLING_API.md - API execution architecture
• SKILL_TEMPLATE.md - Template structure
• agents/README.md - Sub-agent patterns
• skills/limacharlie-call/SKILL.md - Core API skill reference

Plugin root: plugins/lc-essentials/

Step 3: CRITICAL FRAMEWORK RULES

Every skill in lc-essentials MUST follow the LimaCharlie guidelines.

Run /init-lc to load the complete guidelines into your CLAUDE.md. Key rules:

1. Never call MCP tools directly - Use Task with limacharlie-api-executor
2. Never write LCQL queries manually - Use generate_lcql_query first
3. Never generate D&R rules manually - Use AI generation tools
4. Never calculate timestamps manually - Use bash date commands
5. OID is UUID, not org name - Use list_user_orgs to map names to UUIDs
6. Always specify Return field for API executor calls
7. Use sub-agents for parallel operations - Spawn one agent per item
8. Use server-side filtering - selector and online_only parameters

See AUTOINIT.md in the plugin root for complete details.

Step 4: Skill Structure

Create the skill in: skills/{skill-name}/SKILL.md (relative to plugin root)

Required YAML frontmatter:

---
name: skill-name-in-kebab-case
description: Clear description with keywords, use cases, and trigger words. Include action verbs (list, get, create, delete) and domain keywords (sensor, rule, detection). Maximum 1024 characters.
allowed-tools: Task, Read, Bash
---

Model Selection (specified at agent level, not skill level):

• Haiku: Fast, cost-effective for straightforward operations (data gathering, API calls, simple analysis)
• Sonnet: Complex analysis, entity extraction, multi-step reasoning
• Opus: Rarely needed (only for extremely complex tasks)

Step 5: Skill Categories

Determine which category your skill falls into:

Category A: API Wrapper Skill

Simple skills that wrap a single API operation.

• Validate parameters
• Delegate to limacharlie-api-executor
• Format and return results

Category B: Orchestration Skill

Complex skills that coordinate multiple operations.

• Parse user queries
• Fetch required data (e.g., org list)
• Spawn parallel sub-agents (one per item)
• Aggregate and format results

Category C: Research Skill

Skills that search and combine information.

• Search multiple sources with various keywords
• Read multiple files to gather complete info
• Combine information from multiple sources

Step 6: Create Supporting Files (if needed)

If your skill needs a dedicated sub-agent, create it in: agents/{agent-name}.md (relative to plugin root)

Agent frontmatter:

---
name: agent-name
description: What it does and when to use (determines when Claude invokes it)
model: haiku|sonnet|opus
skills:
  - lc-essentials:skill-name
---

Step 7: Update Documentation

After creating the skill:

1. Add entry to SKILLS_SUMMARY.md (in plugin root)
2. If creating an agent, update agents/README.md

Your Task

Based on the user's description above, create a skill that fulfills their requirements.

Workflow:

1. Research Claude Code skills documentation using the claude-code-guide agent
2. Read the plugin framework files listed in Step 2
3. Analyze the user's description to determine:
   • What LimaCharlie API functions are needed
   • Which skill category applies (API Wrapper, Orchestration, or Research)
   • Whether a dedicated sub-agent is needed
   • An appropriate skill name (kebab-case)
4. Ask clarifying questions if the description is ambiguous
5. Create the skill following ALL rules in Step 3
6. Create any supporting agents if needed
7. Update documentation files (SKILLS_SUMMARY.md, agents/README.md if applicable)
8. Report what was created with a summary of the skill's capabilities

Remember: Every rule in Step 3 is CRITICAL. If the skill violates any of these rules, it WILL break the plugin or produce incorrect results.