elasticsearch-domain-delivery

What problem does it solve?

Elasticsearch changes often fail in production due to mapping explosions, analyzer mismatches, relevance regressions, ingest pipeline errors, reindex/alias cutover gaps, and rollback gaps, especially when evidence is missing.

Core Features & Use Cases

• Stack-specific delivery for search/analytics: coordinates mappings, analyzers, index templates, query DSL, ingest pipelines, reindex cutovers, aliases, shard sizing, and ILM/ISM as one contract with explicit rollback discipline.
• Evidence-driven verification: requires scoped checks (e.g., template simulation, token-level analyzer checks via _analyze, query profile/explain or golden queries) and records artifacts before completion.
• Failure-aware implementation flow: supports a blocked/repair loop when required boundaries, ownership, runtime constraints, approvals, or source evidence are missing.

Use Case: implement a safe vN+1 reindex and atomic alias swap for a mapping/analyzer update, then validate relevance with golden queries and confirm lifecycle/shard/observability expectations before claiming completion.

Quick Start

Ask your AI to use Elasticsearch Domain Delivery to implement a mapping/analyzer change with a vN+1 reindex plan, simulated template evidence, scoped query verification, and an explicit rollback path.