Spark Recipe Shared Inbox Status
Review shared inbox health: open vs.
by readdle · MIT
The skill
Review shared inbox health: open vs. done items, unassigned work, and per-member assignment breakdown.
Recipe: Shared Inbox Status
Review the health of one or more shared inboxes - open items, completed work, unassigned emails, and how assignments are distributed across team members.
Prerequisite: Read the use-spark base skill for command reference and filter syntax.
Access level required: read-only.
Steps
Step 1: Discover shared inboxes
spark accounts
Note each shared inbox address and the team it belongs to.
Step 2: Check open items
For each shared inbox:
spark emails shared@co.com:Inbox --filter "is:shared_inbox_open"
These are active items that still need attention.
Step 3: Check unassigned items
spark emails shared@co.com:Inbox --filter "assigned_to:unassigned"
Unassigned items are falling through the cracks - no one owns them yet.
Step 4: Review per-member assignments
spark team "Team Name"
The assignment summary shows how work is distributed. For a deeper look at a specific member's load:
spark emails shared@co.com:Inbox --filter "assigned_to:alice@co.com"
spark emails shared@co.com:Inbox --filter "assigned_to:bob@co.com"
Step 5: Check completed items
spark emails shared@co.com:Inbox --filter "is:shared_inbox_done"
Review recently completed items for a sense of throughput.
Step 6: Present the status
Report per shared inbox:
- Open: N items awaiting action
- Unassigned: M items with no owner
- Per member: assignment counts (flag anyone with significantly more or fewer)
- Done: K items completed recently
If there are multiple shared inboxes, present each separately, then a combined summary.
Tips
- Run this daily for active shared inboxes, weekly for quieter ones.
- Unassigned items are the most actionable finding - they represent work that nobody is looking at.
- Combine per-member counts with
recipe-team-workloadfor a fuller picture of who has bandwidth. - If a shared inbox has high open count but low unassigned count, work is distributed but not being closed - look for stale assignments.
- Use
spark thread <id>to read any item that looks stale or unusual. - For teams with multiple shared inboxes, the total unassigned count across all inboxes is the key metric.
Use this skill
Connect your agent to the Skill Store MCP server and ask for it by ref: readdle/spark-recipe-shared-inbox-status-41708982555.
Details
- Publisher
- readdle
- Category
- Business › Operations & productivity
- License
- MIT
- Updated
- 2026-10-19
References
More from readdle
- Spark Persona Exec Assistant Executive assistant persona for Spark.
- Spark Persona Founder Founder / CEO persona for Spark.
- Spark Persona Freelancer Freelancer / solo operator persona for Spark.
- Spark Persona Meeting Manager Meeting manager persona for Spark.