Agents & Skills - HLMS Docs

Overview

HLMS uses custom Claude configurations to maintain development quality:

Agents - Specialized Claude configurations for testing and analysis
Skills - Reusable knowledge packages for Claude Code

Files are located in:

E:\Lucas\Personal-Productivity\
├── .claude\
│   ├── agents\          # Testing and analysis agents
│   └── skills\          # Reusable skill packages
│       └── hlms-docs\   # Documentation skill

Testing Agent v2.0

The primary agent for quality assurance.

Capabilities

Run test suites (smoke, diff, risky, full, suite)
Analyze test failures with AI
Calculate confidence scores for fixes
Assess regression risk
Detect coverage gaps

Commands

/testing-agent smoke           # Quick sanity check
/testing-agent full            # All 112 tests
/testing-agent suite B         # Specific suite
/testing-agent analyze [bug]   # AI analysis
/testing-agent risk-report     # Risk assessment

Configuration Files

`testing-agent-v2.md`	Main agent configuration (8KB)
`confidence-scoring.md`	4-factor scoring system (5KB)
`regression-risk-scoring.md`	Risk prediction (8KB)
`pattern-changelog.md`	Known bug patterns (6KB)
`test-dependencies.md`	Test cascade mapping (6KB)
`coverage-gap-detector.md`	Coverage analysis (8KB)
`project-suites.md`	Suite definitions

HLMS Docs Skill

Skill for maintaining documentation.

Location

.claude/skills/hlms-docs/
├── SKILL.md              # Main skill file
├── references/
│   ├── page-inventory.md    # All doc pages
│   ├── glossary-terms.md    # Term definitions
│   └── update-checklist.md  # Update process
└── scripts/
    └── generate_docs.py     # Page generator

When It Triggers

Creating/updating user documentation
Adding features that need docs
Generating glossary entries
After code changes affecting UI

Auto-Update Protocol

When implementing features, Claude Code should append:

DOCS UPDATE REQUIRED:
- Page(s): [affected pages]
- Change: [description]
- Glossary: [new terms or "none"]

Using Agents

In Claude Chat (Opus/Sonnet)

Reference agent files for analysis:

"Using the testing-agent-v2.md configuration, 
analyze this test failure..."

In Claude Code (VSCode)

Skills are automatically loaded when relevant. Agent files can be referenced:

"Read .claude/agents/testing-agent-v2.md and 
run a smoke test on the recent changes."

Creating New Agents

Agent files follow this structure:

# Agent Name
## Purpose
[What this agent does]

## Triggers
[When to use this agent]

## Commands
[Available commands]

## Process
[Step-by-step workflow]

## Configuration
[Settings and parameters]

Place in .claude/agents/ directory.

Creating New Skills

Skills follow the standard Anthropic format:

skill-name/
├── SKILL.md           # Required - main skill file
├── references/        # Optional - reference docs
├── scripts/           # Optional - helper scripts
└── assets/            # Optional - templates, images

SKILL.md Structure

---
name: skill-name
description: When to use this skill...
---

# Skill Name

## Quick Reference
[Essential info]

## Detailed Process
[Step-by-step instructions]

## References
[Links to reference files]

Place in .claude/skills/ directory.

Workflow Integration

┌─────────────────────────────────────────────┐
│              DEVELOPMENT WORKFLOW            │
├─────────────────────────────────────────────┤
│                                             │
│  1. Claude Chat (Opus/Sonnet)              │
│     - Uses agents for analysis             │
│     - Generates conceptual prompts         │
│                                             │
│  2. Claude Code (VSCode)                   │
│     - Reads skills for best practices      │
│     - Implements code changes              │
│     - Triggers doc updates                 │
│                                             │
│  3. Testing Agent                          │
│     - Runs appropriate test suite          │
│     - Analyzes failures                    │
│     - Scores confidence                    │
│                                             │
│  4. Docs Skill                             │
│     - Updates affected pages               │
│     - Adds glossary terms                  │
│     - Maintains search index               │
│                                             │
└─────────────────────────────────────────────┘

🤖 Agents & Skills