refine-ticketlisted

# Refine a Ticket Analyze an existing ticket and produce a structured assessment with a Fibonacci sizing estimate grounded in codebase analysis. ## Tools Needed - `Read`, `Write` — read/write queue items and source code - `Grep`, `Glob` — search codebase for referenced modules, files, and test patterns - `Bash` — run `gh issue view` to fetch GitHub issues - `mcp__atlassian__getJiraIssue` — fetch Jira ticket details - `mcp__atlassian__searchJiraIssuesUsingJql` — find related or blocking tickets ## Input A queue item file in `~/.local/share/engineer-agent/queue/incoming/` with type `ticket-refinement`, containing the ticket content in `## Context`. ## Steps ### 1. Read the Ticket Read the queue item to get the full ticket content from the `## Context` section. Extract `project`, `ticket_key`, and `source` from frontmatter. Read `~/.local/share/engineer-agent/engineer.yaml` to find the project config at `projects.<project>` for codebase path and integration settings. ### 2. Analyze the Codebase Using the project path from config: - `Grep` for modules, classes, functions, or files referenced in the ticket - `Glob` to find similar implementations or related code areas - Check existing test patterns (`Glob` for `*test*`, `*spec*`) in the affected areas - Read key files to understand current architecture in the relevant areas ### 3. Cross-Reference Determine the tracker type for this project: - Read `projects.<project>.tracker` from config - If `tracker` is absent, inf