Skip to main content

Version 2026-05-11

Version Information

API Version: 2026-05-11 API Update: 2026-05-11 Plugin Version Required: 3.0.0.0+ Status: Current Release

2026-05-11 is the URL-visible version — it only changes on breaking revisions. api_update bumps on every non-breaking addition within this version, so read it from /guide.api_update at runtime to decide which features are available on the server you are talking to.

Check compatibility: Version Compatibility Guide

Breaking change from 2025-10-20

2026-05-11 introduces the symmetric calc column declaration rule (T87c). Under the previous version, only the from-side material had its calc-referenced columns auto-fetched; join-side references silently returned 0. From 2026-05-11, both sides behave identically: any material.column reference inside a calc expression triggers fetch + preserve regardless of whether the material is reached via from or join.with. Existing queries that worked are unaffected; queries that were silently returning 0 because they relied on join-side calc will now return correct counts. Bad references are caught at validation time with E_CALC_COLUMN_UNRESOLVED. See the Update History for the full rationale.

This page lists what lives inside version 2026-05-11. If you are here for the why behind QAL, read the philosophy pages first — those are version-independent and apply to every API version.

What is in this version​

  • Materials — the data surfaces you can query. One page per material, each with a hand-crafted sample table so the grain of the data is visible at a glance.
  • API Reference — exact shapes of /guide, /query, authentication, and errors, with worked examples.
  • AI Spec — the subset of this directory served to AI / MCP clients via /guide: a concise instruction README and the two machine-readable YAML specs (materials.yaml, qal-validation.yaml).
  • Update History — the Update Ledger for this version. Every non-breaking addition is logged here against its api_update date.

Before you start​

If you are new here:

  1. Read Get Started with AI for the shape of a first interaction.
  2. Read Why QAL, not SQL? for the design rationale.
  3. Then come back here, pick the material you need from the list above, and compose a query.