Skip to content

Triage Workflow Example

Walk through classifying incoming items and managing the backlog. Triage is the first step for any new request, turning unstructured feedback into actionable work.

Scenario

Your team receives support tickets, feature requests, and bug reports throughout the day. AgilePlus triages them automatically into the backlog, prioritizing by type and urgency.

Incoming Items

Bug Report (High Priority)

A customer files a critical bug:

bash
agileplus triage "users can't log in after password reset - getting 500 error"
Analyzing input: "users can't log in after password reset - getting 500 error"

Intent Detection:
  - Primary: Bug (92% confidence)
  - Keywords: error, can't, 500
  - Severity signal: error code + user-facing

Classification
  Type: Bug
  Severity: High
  Priority Score: 95/100

Added to backlog as item #1 [Bug] (Priority: High)

The triage system detects:

  • Type: Bug (broken functionality)
  • Severity: High (login is critical, affects multiple users)
  • Urgency: Critical (prevents user access)

Feature Request (Medium Priority)

bash
agileplus triage "would be great to have dark mode in the dashboard"
Analyzing input: "would be great to have dark mode in the dashboard"

Intent Detection:
  - Primary: Feature (78% confidence)
  - Secondary: Enhancement (15%)
  - Keywords: feature, have, dashboard
  - Qualifier: "would be great" indicates nice-to-have, not blocking

Classification
  Type: Feature
  Scope: UI enhancement
  Priority Score: 58/100

Added to backlog as item #2 [Feature] (Priority: Medium)

Vague Idea (Low Priority)

bash
agileplus triage "what if we integrated with Slack for notifications"
Analyzing input: "what if we integrated with Slack for notifications"

Intent Detection:
  - Primary: Idea (81% confidence)
  - Secondary: Enhancement (12%)
  - Keywords: what if, integrated, Slack
  - Signal: Speculative language ("what if") indicates early-stage thinking

Classification
  Type: Idea
  Scope: Third-party integration
  Priority Score: 35/100

Added to backlog as item #3 [Idea] (Priority: Low)
  → Requires stakeholder validation before planning

Maintenance Task (Medium Priority)

bash
agileplus triage "need to update Node from 18 to 22"
Analyzing input: "need to update Node from 18 to 22"

Intent Detection:
  - Primary: Task (88% confidence)
  - Keywords: update, version bump, maintenance
  - Type signal: Technical maintenance work

Classification
  Type: Task
  Category: Maintenance
  Priority Score: 62/100

Added to backlog as item #4 [Task] (Priority: Medium)
  → Technical debt, affects performance

Duplicate Detection

bash
agileplus triage "dashboard dark mode is needed"
Analyzing input: "dashboard dark mode is needed"

Possible duplicate detected:
  Similar to item #2: "would be great to have dark mode in the dashboard"

  Similarity: 94%

Merge options:
  1. agileplus queue merge 2 --with-new  (combines into #2)
  2. agileplus queue add               (keeps separate)

What would you like to do? [merge/add]: merge

✓ Merged into item #2
  Added note: "Also requested by [user]"

Review the Queue

View all incoming items by priority:

bash
agileplus queue list
Backlog Queue (sorted by priority)

#1  [Bug]     users can't log in after password reset        Priority: HIGH    Status: new
    Severity: Critical  |  Submitted: 2 hours ago  |  Affects: users

#2  [Feature] dark mode in the dashboard                     Priority: MEDIUM  Status: new
    Scope: UI enhancement  |  Submitted: 4 hours ago  |  Votes: 3

#4  [Task]    update Node from 18 to 22                      Priority: MEDIUM  Status: new
    Category: Maintenance  |  Submitted: 1 day ago  |  Blocks: performance work

#3  [Idea]    integrate with Slack for notifications         Priority: LOW     Status: new
    Scope: Integration  |  Submitted: 2 days ago  |  Requires: validation

Filter by Type

bash
# Only bugs
agileplus queue list --type bug

# Open bugs and features
agileplus queue list --type bug --type feature --status new

# High-priority items
agileplus queue list --priority high --priority critical

Get Details on a Queue Item

bash
agileplus queue show 1
Item #1: Login 500 Error After Password Reset

Type: Bug
Status: new
Priority: HIGH (95/100)
Severity: CRITICAL
Submitted: 2 hours ago
Source: customer support ticket #CS-4821

Description:
  "users can't log in after password reset - getting 500 error"

Analysis:
  - Affects: Login flow, password reset flow
  - Impact: Critical (blocks user access)
  - Reproducibility: Confirmed in production
  - Workaround: None known

Related Items:
  - None

Suggested Action:
  → Immediate spec creation + emergency release track
  → Estimated effort: 1 WP, ~1 day

Work the Queue

Pop High-Priority Item

bash
agileplus queue pop
Popped item #1: [Bug] users can't log in after password reset

Priority: HIGH (95/100)
Effort estimate: 1–2 work packages
Time to fix: ~1 day

Ready to specify? Type 'agileplus specify' with title and description.

Convert to Specification

Once triaged, move to planning:

bash
agileplus specify \
  --title "Fix: Login fails with 500 error after password reset" \
  --description "Investigate and fix root cause of 500 error when users attempt to log in immediately after resetting their password. Affects all users." \
  --queue-item 1
Creating spec from queue item #1...

Spec created: 002-login-500-fix
  kitty-specs/002-login-500-fix/spec.md

Queue item #1 marked as: in_progress

You can now plan and implement:
  agileplus plan 002
  agileplus implement 002

Batch Processing

Triage Multiple Items at Once

Process a batch from email, Slack, or support tickets:

bash
# Create a file: incoming.txt
cat > /tmp/incoming.txt <<EOF
Bug: dashboard crashes on large dataset export
Feature: add export to PDF format
Task: migrate database to new schema
Idea: AI-powered test generation
EOF

agileplus triage --batch /tmp/incoming.txt
Processing batch (4 items)...

✓ [Bug]     dashboard crashes on large dataset export      Priority: HIGH
✓ [Feature] add export to PDF format                       Priority: MEDIUM
✓ [Task]    migrate database to new schema                 Priority: MEDIUM
✓ [Idea]    AI-powered test generation                     Priority: LOW

Added 4 items to backlog

Override Classification

If the automatic classifier gets it wrong, override it:

bash
agileplus triage "refactor the auth module to be more testable" --type task
Input: "refactor the auth module to be more testable"

Auto-detected: Feature (65% confidence)
Your override: Task

Added to backlog as item #5 [Task] (Priority: Low)
  Note: Classification was overridden (auto: Feature → actual: Task)

Use --type to specify exactly:

  • bug — broken functionality
  • feature — new capability
  • task — maintenance, refactoring, or housekeeping
  • idea — speculative, needs validation

Weekly Queue Review

Run a weekly status check to monitor backlog health:

bash
agileplus queue health
Backlog Health Report (week of Feb 24 – Mar 2)

Items Added: 12
Items Completed: 4
Items In Progress: 3
Items Waiting: 5

Age Analysis:
  New (< 24h):         3 items
  Recent (1–7 days):   6 items
  Aging (7–30 days):   2 items
  Stale (> 30 days):   1 item ⚠

Priority Distribution:
  Critical: 1 item (8%)
  High: 2 items (17%)
  Medium: 6 items (50%)
  Low: 3 items (25%)

Recommendations:
  → 1 stale item needs closure decision
  → 5 items waiting for clarification
  → Process 2 high-priority items this week

Churn Rate: 33% (items completed vs. added)
Avg. time to spec: 2.3 days
Avg. time in progress: 4.1 days

Integration with Issue Trackers

Sync with GitHub, Plane, or other tools:

bash
# Pull new issues from GitHub
agileplus sync github --direction pull

# This automatically triages GitHub issues into the queue
# Labels on GitHub map to priority and type:
#   type: bug       → Type: Bug
#   priority: high  → Priority: HIGH
Syncing with GitHub...

Fetched 7 new issues
  5 already in queue (duplicate check passed)
  2 new items added to queue

Added to backlog:
  #6  [Bug]     timeout on large API requests
  #7  [Feature] webhook support for webhooks

Key Takeaways

  1. Early classification — Triage immediately when items arrive
  2. Duplicate detection — Merge similar items to avoid work bifurcation
  3. Priority scoring — Automatic detection of severity and urgency
  4. Queue health — Monitor aging items and bottlenecks
  5. Integration — Pull from issue trackers automatically
  6. Override when needed — Auto-classification is good, but human judgment wins

MIT License