Ticket Execution Workflow
**Target Ticket:** $ARGUMENTS
Overview
Ticket Execution Workflow
Target Ticket: $ARGUMENTS
Phase 0: Context Retrieval
-
Fetch the Jira ticket using
{Target_Ticket} -
Extract from ticket description:
{Overview_Document}- Link to the planning/overview document{Implementation_Plan}- Link to the implementation plan document
-
FAILURE CONDITION - Missing or Unclear Links:
If links cannot be found, are ambiguous, or you are unsure which links are correct:
STOP and prompt the user:
I was unable to confidently identify the required documents from ticket {Target_Ticket}. Found links: [list any links found, or "None"] Please provide: 1. Overview Document URL: [paste link] 2. Implementation Plan URL: [paste link]DO NOT PROCEED until both URLs are confirmed.
-
MANDATORY CHECKPOINT - Document Confirmation:
Before proceeding to Phase 1, you MUST confirm with the user:
Please confirm these are the correct documents for {Target_Ticket}: Overview Document: {Overview_Document} Implementation Plan: {Implementation_Plan} Proceed with these documents? (Yes / No / Correct)- Yes → Continue to Phase 1
- No → Ask user to provide correct URLs
- Correct → User provides corrected URLs, then re-confirm
DO NOT PROCEED to Phase 1 without explicit user confirmation.
Phase 1: Preparation
- Read the overview document (
{Overview_Document}) for epic context and requirements - Read the implementation plan (
{Implementation_Plan}) for detailed execution steps - Locate ticket
{Target_Ticket}in the implementation plan - Explore the codebase for relevant patterns before starting
- Check for parallel execution opportunities (see below)
Parallel Execution Check
After reading the implementation plan, check for a Parallel Execution Strategy section:
-
If parallel waves are defined:
- Identify the current wave (first wave with incomplete tickets)
- All tickets in the same wave can be executed simultaneously
- Consider delegating wave tickets to specialized agents in parallel:
- Frontend work → frontend-developer agent
- Backend work → backend-developer agent
- Database work → database-optimizer agent
- Test coverage → test-automator agent
- Present the parallel execution opportunity to the user for approval
-
Parallel execution prompt to user:
The implementation plan defines parallel execution waves. Current Wave: [N] Tickets ready for parallel execution: - TICKET-101: [title] → recommended: [agent-type] - TICKET-102: [title] → recommended: [agent-type] - TICKET-103: [title] → recommended: [agent-type] Options: A) Execute all wave tickets in parallel (spawn multiple agents) B) Execute tickets sequentially (one at a time) C) Select specific tickets to parallelize Which approach? [A/B/C] -
If user approves parallel execution:
- Launch specialized agents concurrently for each ticket
- Each agent works independently on their assigned ticket
- Coordinate completion before advancing to next wave
Phase 2: Implementation
- CHECKPOINT: Confirm ticket selection and approach with user
- Update Jira status to "In Progress"
- Execute implementation steps from the plan
- Write required tests (unit, integration, E2E as needed)
- Verify all acceptance criteria are met
Phase 3: Completion
- Present implementation summary and commit message for user review
Note: The user will handle git commit and PR creation.
User Checkpoints
STOP and check in with the user at these points:
- Before starting: Confirm ticket selection and planned approach
- On deviation: Stop and get approval (see Deviation Handling)
- On blocker: Report the issue and wait for guidance
- When complete: Present commit message and summary for user review
Execution Checklist
Before Starting
- Read the ticket's section in the implementation plan
- Verify all blocking tickets are complete (check Jira status)
- CHECKPOINT: Confirm ticket and approach with user
- Move ticket to "In Progress" in Jira
During Implementation
- Follow implementation steps exactly as documented in the plan
- Reference existing codebase patterns
- Create/modify files as specified
- Write all required tests:
- Unit tests for new functions/components
- Integration tests for API/service interactions
- E2E tests for user flows (if applicable)
Before Marking Complete
- Pre-commit hooks pass (linting, formatting, type checks)
- Pre-push hooks pass (tests, build)
- All new tests written and passing
- All existing tests continue to pass
- 90% branch coverage maintained
- All Jira ticket acceptance criteria met
Completion
- Present commit message and implementation summary for user review
Suggested Commit Message Format
When implementation is complete, present this for user review:
{Target_Ticket}: Brief summary of changes
- Change 1 description
- Change 2 description
- Change 3 description
Deviation Handling
If the implementation plan needs to be deviated from:
-
STOP - Do not proceed with the deviation without approval
-
Report - Clearly explain:
- What the planned approach was
- What deviation is needed
- Why the deviation is necessary
- Impact of the deviation
-
Present options:
- Proceed with deviation (will be documented)
- Modify the approach (suggest alternatives)
- Pause and update implementation plan first
-
WAIT for user feedback before continuing
-
Document the approved deviation in the completion summary
Follow-up Ticket Management
If follow-up work is discovered during implementation:
-
Draft the ticket:
Title: [Clear, actionable title] Description: [What needs to be done and why] Acceptance Criteria: - [ ] Criterion 1 - [ ] Criterion 2 Story Points: [Estimate] Related Ticket: {Target_Ticket} -
Recommend placement:
- Backlog - Low priority, not time-sensitive
- Epic - Related to current epic, schedule during planning
- Current Sprint - Blocking or urgent, needs immediate attention
- Future Sprint - Important but not urgent, schedule next
-
Prompt user: "I've identified follow-up work needed. Here's the draft ticket: [ticket details]
Recommended placement: [placement]
Do you approve? (Yes / No / Modify)"
-
WAIT for approval before creating the ticket in Jira
Output
When complete, provide a summary:
## Ticket Completion Summary
### Ticket
- **Key:** {Target_Ticket}
- **Title:** [Title]
### Suggested Commit Message
[Formatted commit message]
### Changes Made
- [List of changes]
### Files Modified/Created
- `path/to/file.ts` - Description of changes
### Tests Added
- `path/to/test.spec.ts` - What it tests
### Deviations from Implementation Plan
- [Any approved deviations with justification, or "None"]
### Follow-up Tickets Created
- [TICKET-KEY] - [Title] (Placement: [where])
- Or "None"
Agent Delegation
Pre-Implementation Exploration: Before writing code, explore:
- File structure and naming conventions
- Similar existing implementations to reference
- Test patterns and fixtures available
- API patterns and data models
Code Quality: Have implementations reviewed for quality, patterns, and potential issues before committing.
Testing: Delegate test creation to ensure comprehensive coverage.
Test Strategy: Design comprehensive test coverage including:
- Unit tests for business logic
- Integration tests for API/service interactions
- E2E tests for critical user journeys
Quality Validation: Have test coverage and quality reviewed.
Code Review: Have code reviewed for:
- Quality and adherence to patterns
- Security vulnerabilities
- Performance implications
Security Analysis: For sensitive changes, analyze security implications.
Root Cause Analysis: For issues, systematically investigate:
- Error patterns and stack traces
- Recent changes that may have introduced bugs
- Environmental factors
Performance Investigation: For performance issues, profile and analyze bottlenecks systematically.