PromQL Query Formatter & Validator

Format and Validate Prometheus PromQL Queries Instantly

The PromQL Query Formatter & Validator tidies up the queries you write against Prometheus and other compatible time-series databases. Paste in a metric expression and the tool normalises spacing around binary operators, lays out aggregation, function calls and label matchers on their own lines when the query gets long, and points out unbalanced brackets, mismatched braces, missing durations and other syntactic problems before you ever paste the query into a dashboard or alerting rule. Everything runs locally in your browser, so the queries — which often expose internal metric names and labels — never leave your device.

Comment utiliser

1. Paste your PromQL query into the input area, or upload a .promql or text file using the file picker.
2. Pick an indent width — 2 spaces, 4 spaces, or a literal tab character — for any multi-line expansion.
3. Set a maximum line length. Function calls, label matcher sets and low-precedence boolean operators (and, or, unless) wrap onto separate lines once the formatted line would exceed this width.
4. Basculer Always break if you want every aggregation, function-call argument list and label matcher set on its own line regardless of length.
5. Basculer Wrap as recording rule and supply a metric name (such as job:http_errors:rate5m) to emit the formatted query as a ready-to-paste Prometheus recording rule in YAML.
6. Copy the formatted output or download it as a .promql fichier.

Caractéristiques

• PromQL-aware spacing – Inserts a single space around arithmetic, comparison and boolean operators while keeping label matchers tight inside { } braces.
• Smart multi-line layout – Breaks long queries at and / or / unless first, then at function argument boundaries, then at individual label matchers so the result remains easy to scan.
• Aggregation and vector-matching support – Recognises sum, avg, topk, quantile and friends together with by, without, on, ignoring, group_left et group_right modifiers.
• Duration and subquery tokeniser – Parses compound ranges like 1h30m and subqueries like [5m:1m] as single units so they are never split mid-token.
• Validation de syntaxe – Detects unbalanced parentheses, brackets and braces, mismatched openers and closers, invalid label-matcher bodies, invalid subscript ranges and unterminated strings, each reported with a line number.
• Recording rule export – Optionally wraps the formatted query as a Prometheus YAML recording rule block, using a YAML literal block scalar for multi-line expressions so indentation is preserved.

Cas d'utilisation courants

• Cleaning up ad-hoc Prometheus queries copied from Grafana panel editors before committing them to dashboards or alert rules.
• Preparing PromQL expressions for code review — long error-rate or saturation queries become readable diffs once each branch sits on its own line.
• Catching subtle syntax mistakes (missing duration, stray comma, unclosed { after editing label matchers) without round-tripping through a live Prometheus instance.
• Converting raw queries into Prometheus recording-rule YAML for pre-aggregation pipelines.
• Teaching PromQL — the normalised output makes operator precedence and aggregation grouping immediately obvious to readers learning the language.