v0.24.0

SQL Sage 0.24.0 - safer AI agents: Codex now runs read-only without its own shell or web search, and Claude Code turns can use only SQL Sage tools (no file, shell or web access of their own); the AI stops after 12 steps or 10 minutes per answer and tells you so (ask it to continue); new Copy diagnostics button builds a support report you preview first - no SQL, results, names or keys; public changelog on the website.

v0.23.0

SQL Sage 0.23.0 - one-click Explain error: when a query fails, SQL Sage shows what the catalog proves (missing column and the closest real name, the foreign key or CHECK behind a 547, the key columns behind a duplicate, the column size behind a truncation, the parser position of a syntax error) and, with a license, explains and fixes it with AI; values from your rows are never sent; a first-run Check this database (about 10 seconds, read-only) with the top findings and their numbers - free for everyone; subscription licences now always carry an expiry date and are refreshed online.

v0.22.0

SQL Sage 0.22.0 - choose how hard the AI thinks: a reasoning-effort selector next to the model picker (only the levels your model supports) and /effort for a single answer; the model list now comes live from your Claude Code installation; security hardening from an external audit: queries that read from linked servers, files or external sources always ask for confirmation, results you agreed to share with one AI provider are never re-sent to another without asking, Run only executes in the window SQL Sage opened, a more truthful audit trail, chat history encrypted on disk with a retention setting; fixes for chats or probes that could hang when the AI CLI left a helper process running.

v0.21.0

SQL Sage 0.21.0

Free after the trial

Every install still starts with a 30-day trial of everything. When the trial ends, SQL Sage now keeps working in a free mode (SQL Sage Free) instead of switching off. These tools make zero AI calls:

• F5 execution guard — warns before running UPDATE / DELETE without WHERE, DROP and TRUNCATE
• JOIN…ON suggestions built from your foreign keys
• Expand wildcards (SELECT * → column list)
• Qualify object names
• Format SQL — uses SSMS's built-in formatter when it is available, otherwise SQL Sage's own
• Execute current statement — runs just the statement under the caret

AI features (chat, Explain / Fix / Optimize / Document, AI completions) and the proof and audit tools (Change-Impact, Incident Mode, Prove-It) need a license.

SQL Prompt–compatible shortcuts

Shortcut	Command
Ctrl+B, Ctrl+W	Expand wildcards
Ctrl+B, Ctrl+Q	Qualify object names
Ctrl+K, Ctrl+Y	Format SQL
Shift+F5	Execute current statement
Shift+Alt+Z / Shift+Alt+R / Ctrl+Alt+Z	Explain / Fix / Optimize (AI, licensed)

If Redgate SQL Prompt is installed, these shortcuts step aside automatically so they don't clash with it. The commands are always available from the SQL Sage Tools menu when you right-click in the query editor.

"Show impact (Pro)" in the F5 guard

When the guard stops a risky batch, a new Show impact (Pro) link opens the SQL Sage panel with a change-impact request for that SQL (trial or license). Nothing is run. On the free plan, the link explains what a license adds.

Installer fix

The installer is now signed end to end. That includes the setup stage Inno Setup extracts to %TEMP% and the uninstaller. This fixes "Unable to execute file in the temporary directory. Setup aborted. Error 4551" on PCs that use Smart App Control or App Control for Business. Earlier installers had an unsigned inner stage.

Install / upgrade

1. Close SSMS.
2. Run SqlSageSetup.exe (below, or from the download page).

SHA-256 of SqlSageSetup.exe: B6AB0186D18D1F25D5721ADBEC11D8CE39CF8FA9B41ECF74C382E496B5B36E4F

Requirements: SQL Server Management Studio 22 on Windows (x64 / Arm64). AI features use your own Claude (Claude Code) or ChatGPT / OpenAI Codex (Codex CLI) account, so there is no API key. The free tools don't need any AI account.

v0.20.5

SQL Sage 0.20.5

v0.20.4

SQL Sage 0.20.4

v0.20.3

SQL Sage 0.20.3 - privacy fix: without your consent to share results, the AI no longer runs queries on your tables (it sees only column names and types), literals and parameter values in captured query text and plans are masked, and consent is tied to the exact server and database; plus a Buy license button in the trial banner with a gentle reminder in the last 10 days.

v0.20.2

SQL Sage 0.20.2 - sharper, better-grounded answers: longer replies no longer cut off on thinking models (API key), prompt caching for API-key chat, inline completions that never spend their budget thinking, tool descriptions that state their real limits, and no invented table names in completion prompts.

v0.20.1

SQL Sage 0.20.1 - more robust model selection timing: a slow or unresponsive source (the local model catalog, the CLI version check or the model policy) can no longer hold up choosing a model beyond its time budget.

v0.20.0

SQL Sage 0.20.0 - models now follow the CLIs installed on your machine: the Codex channel reads Codex's own local model catalog, so model retirements (such as GPT-5.5 on Oct 14) are handled without an extension update; a clear 'update Codex CLI' message instead of an error when a newer model needs it; never switches you to a costlier model on its own. New 'Recommended' row in the model picker, reasoning effort shown for Codex, inline completions choose their model automatically. Audit records now name the exact model that acted. Hardened validation of model names passed to the CLIs.