Agent Skills: File-Based Todo Tracking

File-Based Todo Tracking

Provides workflows for creating, managing, and completing todos stored as markdown files.

Overview

The todos/ directory contains markdown files with YAML frontmatter for tracking work items.

File Naming Convention

{issue_id}-{status}-{priority}-{description}.md

Examples:
001-pending-p1-security-fix.md
002-ready-p2-performance-opt.md
003-complete-p3-cleanup.md

Status Lifecycle

pending → ready → complete
   ↓
(deleted if skipped)

What do you want to do?

1. Create a todo → See "Creating Todos" below
2. Triage pending items → Use /triage
3. Work on todos → Use /resolve_todo
4. Check dependencies → See "Dependency Management" below

Instrumentation

# Log usage when using this skill
./scripts/log-skill.sh "file-todos" "manual" "$$"

Creating Todos

# Get next ID
next_id=$(ls todos/*.md 2>/dev/null | grep -o '[0-9]\+' | sort -n | tail -1 | awk '{printf "%03d", $1+1}')
[ -z "$next_id" ] && next_id="001"

# Copy template
cp todos/todo-template.md todos/${next_id}-pending-{priority}-{description}.md

Dependency Management

# In YAML frontmatter
dependencies: ["001", "002"]  # Blocked by these issues
dependencies: []               # No blockers

Check blockers:

grep "dependencies:" todos/{file}.md

Quick Commands

# List by status
ls todos/*-pending-*.md
ls todos/*-ready-*.md
ls todos/*-complete-*.md

# List by priority
ls todos/*-p1-*.md

# Count by status
for s in pending ready complete; do
  echo "$s: $(ls todos/*-$s-*.md 2>/dev/null | wc -l)"
done