ebuilder-sql-query-post-pipeline

name	ebuilder-sql-query-post-pipeline
description	Post-pipeline playbook for query-* definitions in configs/sql.*.yml. USE FOR: query.<name>.post task-only result transformation, entries chaining, and baseCommand key-merge override behavior. DO NOT USE FOR: pre-step setup logic.

eBuilder Query Post Pipeline

Purpose

Use this skill when designing post result-processing for query-*.

Scope

• query.<query_name>.post behavior.
• Step chaining through entries.
• Base-command merge/override semantics for post steps.

Current Engine Support

• Post step supports task only.
• Each post step receives exactly one input param: entries.
• The task output (taskResult.Body) is passed as entries to the next step.
• The last post step output is the final query response.

Current limitation:

• query and params fields are not supported in post steps.

BaseCommand Merge Rules

• If only base query defines post, base post is used.
• If both base and query define post, steps are merged by key.
• Same key in query overrides base step.
• Query-only keys are appended after merged base steps.

Example

query:
  query-users-base:
    datasource: main
    command: |
      SELECT id AS "id", name AS "name"
      FROM users
    post:
      - key: normalize
        task: task-normalize-users
      - key: attach-meta
        task: task-attach-base-metadata

  query-users:
    baseCommand: query-users-base
    post:
      - key: normalize
        task: task-normalize-users-v2
      - key: decorate
        task: task-decorate-users

Deterministic Checks

1. Every post step has a unique key.
2. Every post step uses task only.
3. Task contracts accept and return entries shape consistently.
4. Base override keys are intentional and documented.
5. Unsupported post fields (query, params) are not used.