Explained, ready-to-run workflows that chain several tools. Run one here to see what it does, then open it in Pipelines to change its steps, options or input.
Accept sloppy JSON5, sort its keys and ship it compact.
Config files written by hand drift: comments, trailing commas, random key order. This normalises them into canonical, diff-friendly minified JSON.
Pull a sub-document out of an API response and encode it for a header.
Many APIs expect a Base64 JSON blob in a header (e.g. X-Context). jq picks exactly the part you need.
Turn a CSV export into typed JSON, then into YAML you can commit.
Feature flags and environment matrices often live in a spreadsheet. Typed parsing keeps numbers and booleans as real values in the YAML.
Infer column types, generate CREATE TABLE and INSERTs, then pretty-print it.
Seeding a dev database from a CSV export is a daily chore; inferring types saves writing the DDL by hand.
Aggregate a CSV with real SQL, then frame the result as a box-drawn table.
Quick reports for a README, a ticket or a terminal — no database, no spreadsheet.
Select the records with jq and flatten nested fields into dot-path columns.
Analysts want a spreadsheet; the API gives nested JSON wrapped in metadata.
Percent-decode a query value, Base64-decode it and pretty-print the JSON inside.
State, redirect and tracking parameters often hide JSON behind two layers of encoding.
Mask PII and credentials in logs, then run the secret scanner over the result.
Before attaching logs to a ticket or sharing them with a vendor, redact — and double-check with a second, independent scanner.
Sort keys, minify, then SHA-256 — so equal documents hash equally.
Cache keys, idempotency keys and change detection need a hash that ignores key order and whitespace.
Inline the bound parameters from a Hibernate log, then format the SQL.
To reproduce a slow or failing query in a SQL client you need the literal statement, not placeholders.
Convert XML to JSON, pick the items with jq, then flatten to CSV.
Legacy systems speak XML; your spreadsheet doesn't.
Parse YAML to JSON, then validate it with Ajv against your schema.
Catch a typo'd key or a string where a number belongs before the deploy does.
Parse TOML, then query it with jq like any JSON document.
Auditing dependencies across repos is easier once config is JSON.
Turn an indented list into a tree, then put a titled box around it.
Project structures for READMEs and design docs, without drawing a single line by hand.
Slugify each line, drop duplicates and sort.
Generating routes or anchors from a list of titles.
Describe a schema once, generate reproducible rows, emit a Postgres seed script.
Realistic, seeded test data for local development and demos — the same rows every time you run it.
Fill a JSON template with realistic values, flatten to CSV, then profile every column.
Check that a fixture has the spread you expect — nulls, ranges, distinct values — before you build tests on it.
Extract IPv4 addresses with a regex, drop duplicates and sort them naturally.
The first question in any incident: who was hitting us?
Convert a dotenv file to a ConfigMap manifest, then lint the YAML.
Moving an app from docker-compose to Kubernetes starts with its environment.
Call the built-in mock API, then flatten the JSON response to CSV.
Prototype a data export against a fake backend before the real one exists. Edit the routes in Fake JSON API.
Generate Markdown placeholder copy, then render it to HTML.
Fill a layout with realistic headings, paragraphs and lists while the real copy is being written.