add

Add Implementation Work Item to Queue

Add a new implementation work item to the backlog for autonomous processing.

Arguments

• $ARGUMENTS - Brief description of the work item

When to Use /queue:research Instead

Use /queue:research instead of /queue:add when:

• The task requires investigation before coding
• Multiple approaches need evaluation
• Architecture decisions need documenting
• Scope is unclear and needs a plan first

/queue:add is for known work — you know what needs to be built and how.

Process

1. Read .claude/queue/backlog.json to get current items

2. Gather required information:

   • Title: From the arguments (or ask if unclear)
   • Priority: 1 (urgent) to 5 (low) - ask if not specified
   • Owner: Which agent should handle this (e.g., staff-engineer, staff-software-architect)
   • Problem: What needs to be solved
   • Acceptance criteria: How do we know it's done

3. Verify the task fits story sizing rules:

   • Can it be completed in a single context window?
   • Does it touch 3-5 files or fewer?
   • If not, suggest splitting into smaller tasks

4. Create the work item:

   {
     "id": "YYYY-MM-DD-{slug}",
     "title": "Brief description",
     "type": "implementation",
     "priority": 1-5,
     "status": "pending",
     "requires_approval": false,
     "owner": "agent-type",
     "blocked_by": [],
     "context": {
       "problem": "What needs solving",
       "acceptance": ["Criterion 1", "Criterion 2"],
       "files_hint": ["optional/path/hints"]
     },
     "sizing": "single-context",
     "created": "ISO timestamp",
     "updated": "ISO timestamp"
   }
   

5. Check for dependencies:

   • If this task depends on another task, add the dependency ID to blocked_by
   • Blocked tasks won't be picked up by /queue:work until dependencies are resolved

6. Add to backlog.json and write the file

Confirmation

After adding, report:

• Work item ID
• Title and priority
• Assigned owner
• Number of acceptance criteria
• Current backlog size
• Any blocking dependencies