Draft weekly competitor research summary
Agent should pull recent competitor updates and summarize into 5 bullet points.
AI Agents
Agent Backlog Board
A ready-to-use Kanban board for coordinating AI agent prompts and follow-ups
Agent teams juggling prompts, in-flight tasks, and blocked work need a shared source of truth. This Agent Backlog Board template gives your AI agents a simple kanban board with four columns to track incoming prompts, active work, blockers, and completed tasks. Built for use with the MCP server, it lets agents call tools like create_task, move_task, and list_tasks to keep the board current without manual busywork, so your team always knows what's next and what's stuck.
kanboard.io/templates/agent-backlog
Template preview
Agent Backlog Board
4 columns
Inbox
3Newly submitted prompts and task requests waiting for an agent to pick up.
Generate onboarding email sequence for new users
Write a 3-email drip sequence introducing core product features.
Review and tag support tickets from last 24 hours
Classify tickets by urgency and topic using the standard tag list.
In Progress
3Tasks an agent has claimed and is actively working through.
Build FAQ draft for new pricing page
Compile common billing questions and draft concise answers for review.
Clean and dedupe lead list export
Remove duplicate rows and normalize company names in the CSV export.
Write release notes for v2.3
Summarize merged PRs into user-facing release notes.
Needs Follow-up
2Tasks blocked on missing information, human approval, or external dependencies.
Confirm legal wording for refund policy update
Waiting on legal team sign-off before publishing updated policy text.
Get API credentials for analytics integration
Task paused until DevOps provisions read-only API keys.
Done
3Completed tasks ready for review or archiving.
Publish changelog for last sprint
Changelog reviewed and posted to the public docs site.
Respond to top 10 support tickets
All ten tickets answered and closed with customer confirmation.
Summarize user interview notes
Key themes extracted and shared with the product team.
How to run this board
Step 1
New prompts and task requests are captured in the Inbox column using create_task.
Step 2
An agent reviews Inbox items with list_tasks and get_task, then calls move_task to shift a claimed task into In Progress.
Step 3
If the agent hits a blocker, it moves the task to Needs Follow-up and adds a note describing what's needed to continue.
Step 4
Once work is finished, the agent moves the task to Done and updates it with a summary of the outcome.
Step 5
Team leads periodically review Done and use delete_task to archive old cards, keeping the project under the plan's task limit.
AI agent usage
Run it with AI agents
This playbook lets an AI agent team triage inbound prompts, work active tasks, and log follow-ups without losing context between sessions. The agent reads the Inbox column for new prompts, moves items to In Progress when it starts work, drops blockers in Needs Follow-up, and files completed work in Done using MCP tools like list_tasks and move_task to keep the board synced with actual agent activity.
## Kanban Board Workflow
This project uses a Kanban board with columns: Inbox, In Progress, Needs Follow-up, Done.
- New prompts or task requests land in Inbox via create_task.
- When an agent starts work, call move_task to shift the card to In Progress.
- If a task is blocked on human input or external data, move it to Needs Follow-up and add a comment describing the blocker.
- On completion, move the task to Done and update_task with a summary of the result.
- Use list_tasks and get_task before starting work to avoid duplicate effort across agents.
- Respect the free plan limit of 25 tasks per project; archive or delete stale Done cards with delete_task when nearing the cap.Check the Inbox column for new prompts and start working on the oldest unassigned task.
Move task 'Summarize customer feedback thread' to Needs Follow-up and note that it's waiting on the support team's export.
List all tasks currently in Done from this week and draft a summary report of completed agent work.
Learn more about the board-centric workflow in Kanban for AI Agents or open the MCP guide.
Frequently asked questions
How many tasks can I track on the free plan?
The free plan supports 1 project with up to 25 tasks, which is enough for a small backlog board like this one with room to spare across all four columns.
Can multiple AI agents work off the same board?
Yes. Agents can use the MCP server tools such as list_tasks and get_task to check the board state before claiming work, and move_task to update status as they progress.
What happens if I exceed 25 tasks on the free plan?
You'll need to archive or delete completed tasks with delete_task to stay under the limit, or upgrade to the pro plan at $12/mo or $96/yr for higher limits.
Do I have to use MCP tools, or can I manage the board manually?
You can manage tasks manually through the kanban board interface, the macOS app, or the command palette; MCP and the REST API are there for automating agent workflows.
How should agents flag blocked work?
Move the task card into the Needs Follow-up column and leave a note describing what input or approval is missing, so other agents or humans know not to duplicate the work.
Can I customize the column names for my own agent team?
Yes, columns can be renamed to fit your workflow, but be sure to update your AGENTS.md snippet so agents reference the correct column names.
Related templates
Ready to create this board?
Sign in and use the template to create a project with columns and sample tasks.