6.4 KiB

Raw Blame History

Step 4: Triage Each Bug

MANDATORY EXECUTION RULES (READ FIRST):

🛑 NEVER generate content without user input
📖 CRITICAL: ALWAYS read the complete step file before taking any action
🔄 CRITICAL: When loading next step with 'C', ensure the entire file is read and understood before proceeding
✅ ALWAYS treat this as collaborative triage between peers
📋 YOU ARE A FACILITATOR - ask clarifying questions when needed
🎯 This step performs the CORE TRIAGE analysis
⚠️ ABSOLUTELY NO TIME ESTIMATES - AI development speed varies widely
✅ YOU MUST ALWAYS SPEAK OUTPUT In your Agent communication style with the config {communication_language}

EXECUTION PROTOCOLS:

🎯 Triage ONE bug at a time with user confirmation
💾 Track triage decisions for bugs.yaml update
📖 Ask clarifying questions when severity/complexity unclear
🚫 FORBIDDEN to auto-triage without user review

CONTEXT BOUNDARIES:

Parsed bugs from step-03 are in memory
Reference bugs.yaml header for severity/complexity definitions
Reference epics.md for story mapping
Each bug gets full triage analysis

YOUR TASK:

Perform collaborative triage analysis on each parsed bug, assessing severity, complexity, effort, workflow routing, and documentation impact.

TRIAGE SEQUENCE (FOR EACH BUG):

1. Generate Bug ID

Find highest existing bug-NNN from step-03 grep results
Assign next sequential ID (e.g., bug-006)
Format: bug- + zero-padded 3-digit number
For features: feature- + zero-padded 3-digit number

2. Assess Severity

Severity Levels:

Level	Criteria
critical	Prevents core functionality, crashes, data loss
high	Blocks major features, significantly degrades UX but has workaround
medium	Affects subset of users, minor impact
low	Cosmetic, edge case, minor inconvenience

Analysis Questions:

Does it prevent core functionality? → critical
Does it cause crashes or data loss? → critical
Does it block major features? → high
Does it significantly degrade UX but have workaround? → high
Does it affect subset of users with minor impact? → medium
Is it cosmetic or edge case? → low

If Unclear - ASK: "Clarification needed for: {bug_title}

I need more information to assess severity:

Does this bug prevent users from completing core flows?
Does the bug cause crashes or data loss?
How many users are affected? (all users, specific platform, edge case)
Is there a workaround available?

Please provide additional context."

3. Assess Complexity

Complexity Levels:

Level	Criteria
trivial	One-line fix, obvious solution
small	Single file/component, solution clear
medium	Multiple files OR requires investigation
complex	Architectural change, affects many areas

If Unclear - ASK: "Clarification needed for: {bug_title}

To estimate complexity, I need:

Have you identified the root cause, or does it need investigation?
Which file(s) or component(s) are affected?
Is this isolated or does it affect multiple parts of the app?

Please provide technical details if available."

4. Determine Workflow Routing

Routing Matrix:

Severity	Complexity	Workflow
critical	any	correct-course
high	trivial	direct-fix
high	small	tech-spec
high	medium/complex	correct-course
medium	trivial	direct-fix
medium	small	tech-spec
medium	medium/complex	correct-course
low	trivial	direct-fix
low	small+	backlog

If bug mentions story ID (e.g., "2-7"), use that
Otherwise, infer from description using epic keywords
Reference epics.md for story matching
Format: {epic_number}-{story_number} or null

6. Determine Affected Platform

Extract from description:

all - Default if not specified
ios - iOS only
android - Android only
web - Web only

7. Assess Documentation Impact

PRD Impact (doc_impact.prd: true/false) Set TRUE if issue:

Conflicts with stated product goals
Requires changing MVP scope
Adds/removes/modifies core functionality
Changes success metrics
Affects multiple epics

Architecture Impact (doc_impact.architecture: true/false) Set TRUE if issue:

Requires new system components
Changes data model (new tables, schema)
Affects API contracts
Introduces new dependencies
Changes auth/security model

UX Impact (doc_impact.ux: true/false) Set TRUE if issue:

Adds new screens or navigation
Changes existing user flows
Requires new UI components
Affects accessibility

If any doc_impact is TRUE AND workflow != correct-course:

Override workflow to correct-course
Add note: "Workflow elevated due to documentation impact"

8. Add Triage Notes

Document reasoning:

Why this severity? (business impact, user impact)
Why this complexity? (investigation needed, files affected)
Why this workflow? (routing logic applied)
Suggested next steps or investigation areas

9. Present Triage for Confirmation

"Triage: {bug_id} - {bug_title}

Field	Value
Severity	{severity}
Complexity	{complexity}
Platform	{platform}
Workflow	{recommended_workflow}
Related	{related_story or 'None'}

Documentation Impact:

PRD: {yes/no}
Architecture: {yes/no}
UX: {yes/no}

Triage Notes: {triage_notes}

[A] Accept triage [M] Modify - adjust severity/complexity/workflow [S] Skip - don't triage this item now [N] Next bug (after accepting)"

10. Handle Modifications

If user selects [M]:

Ask which field to modify
Accept new value
Re-present triage for confirmation

SUCCESS METRICS:

✅ Each bug triaged with user confirmation ✅ Unclear items prompted for clarification ✅ Routing matrix applied correctly ✅ Documentation impact assessed ✅ Triage notes document reasoning

FAILURE MODES:

❌ Auto-triaging without user review ❌ Not asking clarifying questions when needed ❌ Incorrect routing matrix application ❌ Missing documentation impact assessment ❌ Not documenting triage reasoning

❌ CRITICAL: Reading only partial step file ❌ CRITICAL: Proceeding without user confirmation per bug

NEXT STEP:

After ALL bugs triaged (user selected [A] or [N] for each), load ./step-05-update.md to update bugs.yaml and bugs.md.

Remember: Triage each bug individually with user confirmation!

6.4 KiB Raw Blame History