perf(backend): make analytics pagination stable and bounded for large vault histories

[arisu6804]

Problem

Offset pagination over growing transaction histories becomes slow and can produce duplicates or gaps while new records arrive.

Objective

Deliver a production-quality improvement to analytics history queries and API pagination that creates measurable value for correctness, security, reliability, performance, or maintainability.

Implementation scope

• Replace unbounded offsets with a documented cursor or stable snapshot strategy.
• Enforce maximum page size and apply indexed filters for vault, actor, and time range.
• Return deterministic ordering and cursor metadata that survives normal inserts.

Acceptance criteria

• Repeated pages contain no duplicates or gaps under concurrent inserts.
• Requests above the limit are rejected or capped explicitly.
• Query plans and a representative benchmark demonstrate bounded behavior.

Required validation

• Pagination contract tests, concurrent-insert integration tests, and a large-fixture performance check.
• Existing tests and CI remain passing.
• Add regression coverage for the original failure mode.
• Do not weaken, delete, or skip unrelated tests to obtain a green build.

PR quality bar

• Keep the PR focused on this issue and explain design tradeoffs.
• Include test evidence, compatibility impact, and migration or rollout notes where relevant.
• Avoid typo-only, documentation-only, cosmetic-only, or unrelated refactor submissions.

Out of scope

• Broad rewrites not required by the acceptance criteria.
• Changes to unrelated services, contracts, or user flows.