Agent skill

vtm-expert

Virtual Task Manager domain expert. Knows about: - Next, Context, Task, Start, Complete, Stats, List operations Use when: - User asks about task management operations - Context needed before starting work - Reviewing task status - Managing task workflow

Stars 163
Forks 31

Install this agent skill to your Project

npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/data/vtm-expert

SKILL.md

VTM Expert Skill

What This Skill Does

Helps you manage your Virtual Task Manager workflow with smart command suggestions.

Auto-discovers when you mention task-related work and suggests the appropriate VTM command.

Available Commands

  • /vtm:next - Get next ready task to work on
  • /vtm:context - Generate minimal context for a specific task
  • /vtm:task - View details of a specific task
  • /vtm:start - Mark a task as in-progress
  • /vtm:complete - Mark a task as completed
  • /vtm:stats - Show VTM statistics and progress
  • /vtm:list - List all tasks with their status

When Claude Uses This

When you mention things like:

  • "What should I work on?" → Suggests /vtm:next
  • "I need context for TASK-003" → Suggests /vtm:context TASK-003
  • "Show me task status" → Suggests /vtm:stats or /vtm:list
  • "Start working on TASK-005" → Suggests /vtm:start TASK-005
  • "Mark TASK-003 as done" → Suggests /vtm:complete TASK-003
  • "How many tasks are left?" → Suggests /vtm:stats

VTM Workflow

The typical VTM workflow:

  1. Find work: /vtm:next - Shows ready tasks (dependencies met)
  2. Get context: /vtm:context TASK-XXX - Token-efficient task context
  3. Start task: /vtm:start TASK-XXX - Mark as in-progress
  4. Implement: Use PROMPT 2 with TDD based on test_strategy
  5. Complete: /vtm:complete TASK-XXX - Mark as done, unblocks dependents
  6. Repeat: Back to step 1

Token Efficiency

VTM achieves 99% token reduction by:

  • Surgical access to specific tasks (not loading entire manifest)
  • Dependency resolution built-in
  • Two context modes: minimal (~2000 tokens) and compact (~500 tokens)
  • No need to load related specs unless specifically needed

Best Practices

  1. Before starting: Run /vtm:next to see what's ready
  2. Get context first: Always use /vtm:context before implementing
  3. Follow TDD: Respect the task's test_strategy field
  4. Track progress: Use /vtm:stats to monitor overall progress
  5. Update status: Mark tasks in-progress and completed for accurate tracking

Integration

Works seamlessly with other Claude Code domains:

  • Can be invoked manually: /vtm:operation
  • Auto-triggered by Claude based on conversation
  • Integrates with git hooks for validation
  • Supports team workflows via shared vtm.json

Customization

Edit trigger phrases in the frontmatter above to match your vocabulary.

Examples:

  • Add project-specific terms: "backlog item", "sprint task"
  • Add abbreviations: "ctx" for context, "nxt" for next
  • Add workflow terms: "pick up task", "finish task"

Technical Details

Architecture:

  • Commands wrap the vtm CLI (npm package)
  • Requires vtm.json in project root
  • Atomic writes with automatic stats recalculation
  • Dependency validation built-in

Data Flow:

User phrase → Skill triggers → Command suggested →
User approves → vtm CLI executes → Stats auto-update

See Also

  • Design spec: .claude/designs/vtm.json
  • Commands: .claude/commands/vtm/
  • Prompts: prompts/1-generate-vtm.md, prompts/2-execute-task.md, prompts/3-add-feature.md

Expand your agent's capabilities with these related and highly-rated skills.

Didn't find tool you were looking for?

Be as detailed as possible for better results