Skills Reference

LLPM implements the Agent Skills standard for reusable AI instruction sets. This means skills you create for LLPM can be used with other compatible systems like Claude Code.

LLPM includes 19 core skills organized into 6 categories:

Category	Skills	Purpose
Planning	5	Project planning, requirements, timelines
Visualization	3	Architecture diagrams, dependency maps
Communication	2	Stakeholder updates, meeting agendas
Documentation	3	Markdown formatting, user stories, FAQs
Analysis	3	Risk detection, triage, gap analysis
Research	3	Web research, summaries, note synthesis

How Skills Work

Discovery: LLPM scans skill directories for SKILL.md files
Loading: Skills are parsed and registered in the system
Activation: Enabled skills are injected into the AI’s system prompt
Execution: The AI follows skill instructions when relevant

Managing Skills

Use the /skills command to manage skills:

# List all discovered skills
/skills list

# Preview a skill without enabling
/skills test <skill-name>

# Enable a skill for the current session
/skills enable <skill-name>

# Disable a skill
/skills disable <skill-name>

# Rescan directories for new/modified skills
/skills reload

# Reinstall skills from the repository
/skills reinstall

Skill Locations

LLPM discovers skills from multiple locations:

Location	Purpose	Priority
`~/.llpm/skills/`	User-installed skills	Highest
`.skills/`	Project-specific skills	Medium
Built-in	Default LLPM skills	Lowest

Skills with the same name are loaded based on priority (higher priority wins).

Learn More

Creating Skills - How to create custom skills
Planning Skills - Project planning and decomposition
Visualization Skills - Diagrams and visual representations
Communication Skills - Stakeholder updates and meetings
Documentation Skills - Formatting and templates
Analysis Skills - Risk detection and triage
Research Skills - Web research and summarization