Cursor rule for Work Coordination & Task Prep Mode - install the rule and tell the agent "What's next?" or "It's standup time"

work-coordination.mdc

---
description: Work Coordination & Task Prep Mode
globs:
alwaysApply: true
---

# Work Coordination & Task Prep Mode

## Activation Protocol

**This mode activates in TWO ways for maximum reliability:**

### 1. AUTOMATIC (Proactive - No user prompt needed):
- After reading any Jira ticket → Search for related tickets, blockers, team work
- When discovering a blocker → Immediately search Jira for related work
- After completing implementation → Auto-generate status summary
- **Before providing ANY status update → Re-query Jira for fresh data (MANDATORY)**
- When referencing any ticket number → Auto-check current status

### 2. USER-TRIGGERED (Reactive - Keywords activate mode):
- Any mention of "standup" (including "standup time", "it's standup", etc.)
- Any mention of Jira tickets, stories, or issue keys (e.g., "PV-", "check ticket")
- Preparing for standup meeting
- Generating standup report
- Picking up next task
- What to work on next
- Task recommendations
- Sprint planning preparation
- Work coordination
- Blocker status or updates
- Progress on tickets

**Both paths execute the same search sequence below - ensuring no missed coordination.**

## Execution Protocol

Execute this search sequence using Jira MCP tools (`mcp__mcp-atlassian__jira_*` or others available):

### 1. User Assigned Tickets

- DONE in last 7-14 days
- IN PROGRESS items
- BLOCKED items

JQL:
```
assignee = currentUser() AND status CHANGED TO (Done, Closed) AFTER -14d
assignee = currentUser() AND status in ("In Progress","In Review","QA","Blocked")
```

### 2. Parent Epic Scan

For each epic containing user's active work:
- List all active child issues for any team member
- Filter: linked/dependent, same component, or impacts user

### 3. Similarity Search (Top 5)

Search by components/labels/keywords for tickets with similar functionality.
Detect: duplication, conflicts, shared libraries, coordination needs.

### 4. Linked Issues + Dependency Mapping

- Inspect linked issues and blocks/blocked-by relationships
- Check PR/development links; state if unavailable
- Identify blockers affecting sprint, user progress, or recommended next work

### 5. Candidate Pool for Next Tickets

Gather 3-8 candidates from:
- Unassigned or Ready items in relevant epics/components
- User's backlog/To Do

Confirm: Ready/To Do status, not blocked, appropriate for sprint/backlog.

## Decision Rubric

Rank candidates by:
1. Unblocks other tickets/people
2. Supports sprint goal or top priority
3. Ready + small + low dependency risk
4. Reduces risk (aging/stalled/high-severity)
5. Stakeholder urgency/SLA

Produce "Defer/Do Not Pick Up" list with explicit reasons and required changes.

## Output Format

Total: 2-3 minutes reading time. Each section: 30-60 seconds. Brief, actionable, coordination-focused.

```
Sprint Standup & Work Selection

DONE (Last 7-14 days) (3-5 bullets max)
• ticket key + outcome delivered

IN PROGRESS (3-5 bullets max)
• ticket key + current focus + next step

BLOCKERS (1-4 bullets max)
• ticket key + blocker + who/what needed + next action + ETA if known

RELATED TEAM WORK (3-6 bullets max)
• ticket key + owner + relation (same epic/blocks/depends/shared component)

RECOMMENDED NEXT WORK (1-3 bullets max)
• ticket key + why (impact/readiness/unblock benefit)

DEFER / DO NOT PICK UP (1-4 bullets max)
• ticket key + why defer + what must change to pick up
```

If section exceeds cap, group by theme and keep highest-leverage items.

## Anti-Hallucination Enforcement

Use ONLY data from Jira MCP tools. State "No tickets found" if searches return empty. Link to actual ticket keys. Cite actual status, assignee, dates.

Do not invent tickets, statuses, blockers, owners, PRs, dates, estimates, or epic names. Do not assume ticket content without reading.

## Mode Deactivation

Active until user changes topic or asks non-work-coordination question.