## 1. Golden Cases

- [x] 1.1 Create retrieval evaluation directory structure.
- [x] 1.2 Add fixed golden retrieval cases covering Chat and AIOps retrieval scenarios.
- [x] 1.3 Add matching offline retrieval fixtures for every golden case.

## 2. Evaluator

- [x] 2.1 Implement an offline retrieval evaluator script.
- [x] 2.2 Support hit-level classification and first expected document rank.
- [x] 2.3 Support JSON and Markdown report output.

## 3. Baseline Report

- [x] 3.1 Generate the baseline JSON report from the fixed cases and fixtures.
- [x] 3.2 Generate the baseline Markdown report from the fixed cases and fixtures.

## 4. Documentation

- [x] 4.1 Document the retrieval baseline purpose, file layout, and regeneration command.
- [x] 4.2 Link the retrieval baseline from the RAG refactor issue or related MVP documentation.

## 5. Verification

- [x] 5.1 Run the evaluator successfully against the fixed baseline cases.
- [x] 5.2 Run OpenSpec status/validation for the change and confirm tasks are complete.