Bulk-Plan Your Ticket Backlog
Multi-select tickets in Linear and get confidence-scored implementation plans.
Connect the ticket system and write the planning playbook
Connect Linear, Asana, ClickUp, or monday.com so Opulent can read the backlog and post comments. The planning logic lives in a playbook, a reusable, named set of steps that tells Opulent how to scope a ticket without writing code.
The playbook defines the confidence score rubric, the file analysis steps, and the format of the plan comment.
Playbook: !plan For each selected ticket: 1. Read the title, description, and comments. 2. Search the codebase for relevant files and patterns. 3. Identify the change surface and any existing similar code. 4. Score confidence: - Green: change is isolated, pattern exists, no open questions. - Yellow: feasible, but one or more clarifications needed. - Red: too large, unclear, or risky; best for a human. 5. Post a comment with the score, relevant files, implementation plan, and open questions. 6. Do not write code unless explicitly told to.
Seed the playbook with your team's complexity thresholds. A 'green' ticket for a senior engineer might be 'yellow' for a junior, but the confidence score should reflect how well Opulent understands the work, not who will do it.
Select tickets and run the planning pass
Multi-select tickets in Linear and add a label that triggers Opulent, or run the playbook manually against a list of ticket IDs. Opulent scopes every selected ticket in parallel and posts a structured comment on each.
The trigger should target a specific status or label so only tickets that need planning get the treatment. Edge detection prevents re-planning tickets that were already scored.
Trigger: label "!plan" added to a ticket Run for each ticket: - Read ticket and linked context. - Search codebase for relevant files. - Score and plan. - Post comment with score, plan, and questions. Or bulk mode: - Provide a list of Linear IDs. - Opulent runs the same analysis in parallel.
Read the analysis on each ticket
Opulent posts a comment with the confidence score, relevant files, and a step-by-step implementation plan. The score tells the team whether the ticket is ready to hand to Opulent or needs human review first.
Scoring summary: | Ticket | Title | Score | Relevant files / notes | |--------|------------------------------|-------|------------------------| | FE-412 | Dark mode toggle | Green | ThemeContext exists; 3 toggles already use it. | | FE-419 | Date picker timezone bug | Green | Bug isolated to formatLocalDate(); test missing. | | BE-203 | Migrate prefs to Postgres | Yellow| 14 call sites; needs rollback strategy. | | BE-210 | Rate limiting on webhooks | Green | Existing middleware; needs config. | | INF-102| k8s health check | Green | Readiness probe exists; liveness follows. | Comments posted on each ticket.
Set up continuous planning
Configure an automation so new tickets are planned the moment they hit a status like 'Triage' or 'Todo'. Opulent scores them and updates labels, so the team sees at a glance which tickets are ready to implement and which need clarification.
Automation: Trigger: ticket moves to "Todo" in Engineering team Filter: label is "Feature" or "Improvement" Action: start Opulent run with !plan playbook After scoring: - Green label -> ready for !implement - Yellow label -> needs clarification - Red label -> assign to senior engineer
Sharpen the planning loop
When a plan is wrong, update the playbook or write the correction into memory (the notes a run recalls next time). When a ticket type is consistently scored wrong, add an eval that checks the plan against known good examples.
The natural chain: green tickets can move to Auto-Triage Bugs via Linear or Turn PRDs into Sprint Tickets for implementation; yellow tickets can route to Autonomous Triage for Knowledge Work Queues for clarification.