Reports

Warestack turns your operational data — from GitHub, Slack, Linear, and Jira — into structured, queryable reports.
Instead of manually searching logs or PR histories, you can use natural language to ask questions and instantly see what went wrong.

💡 Reports are built from prompts. Warestack converts your query into a deterministic report.

Example query

Imagine your CXTO requests you to extract PRs where files under auth/ or secrets/ changed in the past month. Using Warestack you can automatically surface relevant PRs in the form of structured results that explain who changed what and why.

Example of Warestack's custom check — From your mind to an active check.

Warestack converts your prompt into a deterministic query behind the scenes:

SQL

SELECT pr.id, pr.title, pr.author, pr.merged_at, f.file_path
FROM pull_requests pr
JOIN pr_files f ON pr.id = f.pr_id
WHERE (f.file_path LIKE 'auth/%' OR f.file_path LIKE 'secrets/%')
AND pr.merged_at >= DATE_SUB(CURRENT_DATE, INTERVAL 30 DAY)
ORDER BY pr.merged_at DESC;

Types of Reports

Warestack supports two types of reports:

On-the-fly reports generated from prompts.

These reports are generated on the fly from natural language prompts. Warestack turns your question into an SQL query, lets you review or tweak it, and then runs it instantly on live data.

💡 Ideal for quick insights, ask, tune, and save your report in seconds.

Stored reports with deterministic SQL.

These are saved reports with fixed SQL queries. You can rerun or schedule them for automatic delivery to your team. Ideal for monthly or quarterly reviews with full traceability and zero LLM errors.

💡 Schedule stored reports to deliver recurring analytics or compliance summaries automatically.