Decide Whether to Rewrite the Legacy System or Patch It Again

Another patch went in last week and everyone quietly knows it's the fourth one this quarter. This weighs the real cost of patching again against the real cost of a rewrite, looks for a genuine middle option instead of framing it as all-or-nothing, and lands on a recommendation with a specific trigger for revisiting it.

<context>
You are a seasoned technical director advising on one of the most consequential and most avoided decisions in engineering: whether to finally rewrite a legacy system or patch it once more. {SYSTEM_NAME} is the system in question. [describe its age, its current problems, and what a rewrite would need to replace] is the current situation.
</context>

<task>
**Work through the decision:**
1. Assess the real cost of continuing to patch: rising complexity, slowing delivery, key-person risk, and how much longer the system can realistically absorb further patches
2. Assess the real cost and risk of a rewrite: time, the risk of missing hidden business logic, the cost of running both systems in parallel, and the opportunity cost of the team's time
3. Identify a middle option if one genuinely exists, such as a targeted rewrite of the worst component while leaving the rest alone, rather than framing it as all-or-nothing
4. Recommend a path with a clear, specific trigger condition that would justify revisiting the decision later, for example a second major outage, or a named feature becoming impossible to add
</task>

<output_format>
- A cost and risk assessment of continuing to patch, one paragraph
- A cost and risk assessment of a full rewrite, one paragraph
- A middle option, described specifically, or a clear statement that none genuinely exists
- A final recommendation with a named trigger condition for revisiting it
- Tone: senior and unsentimental, this decision has been avoided long enough
</output_format>