| name | plan-to-hc |
| description | Create a hive hc epic from the current plan file, inserting the ENTIRE PLAN TEXT into the epic and creating subtasks to track all work items.
|
| allowed-tools | Read,Bash(hive hc:*),Bash(echo:*),Bash(bpcopy:*),Bash(cat:*) |
| disable-model-invocation | true |
Plan to Hive HC
Converts a Claude Code plan file into a hive hc epic with subtasks.
Instructions
When this command is invoked:
-
Locate the most recent (active) plan file: Read the current plan file from the plans directory
-
Extract the plan content: Get the ENTIRE TEXT of the plan
-
Create a hive hc epic with subtasks: Use the bulk JSON stdin format to create the epic and all subtasks in a single command. This avoids shell escaping issues with --desc.
Build a JSON object and pipe it to hive hc create:
cat <<'JSONEOF' | hive hc create
{
"title": "Epic title here",
"type": "epic",
"desc": "Full plan text here (escaped for JSON)",
"children": [
{"title": "Step 1: ...", "type": "task"},
{"title": "Step 2: ...", "type": "task", "desc": "details"}
]
}
JSONEOF
- The epic description must contain the COMPLETE plan text
- Each major work item becomes a child task
- Use clear, actionable titles for subtasks
CRITICAL: Do NOT use --desc flag for long text. Always use the JSON stdin format to avoid the description being added as a comment instead.
-
Copy the epic id: Put the epic id on the clipboard
Notes
- ALWAYS use the JSON stdin format — never pass plan text via
--desc flag (it gets added as a comment instead of description)
- The epic description must contain the COMPLETE plan text
- Subtasks should capture all trackable work items from the plan
- Escape the plan text properly for JSON (newlines as
\n, quotes as \")
- Preserve any dependencies or sequential ordering from the plan