Proposal
Measure memory and latency for mocked multi-page collections near the 2,000-item default guard and document a regression budget.
Acceptance criteria
- Fixtures exercise item, page, repeated-link, and truncation guards.
- Results report p50/p95/p99 latency and peak memory.
- CI enforces a stable bounded regression threshold.
- The benchmark never calls the live GitHub API.
Deferred: Outside frozen run scope
This proposal was discovered during the frozen public-issue run and is intentionally excluded from implementation in that run.
Proposal
Measure memory and latency for mocked multi-page collections near the 2,000-item default guard and document a regression budget.
Acceptance criteria
Deferred: Outside frozen run scope
This proposal was discovered during the frozen public-issue run and is intentionally excluded from implementation in that run.