aq

aq is a jq-style CLI for querying and rewriting structured data across more than just JSON.

It currently supports:

• jq-style query semantics and result streams
• JSON, JSONL, YAML, TOML, CSV, and TSV input and output
• output-only table rendering for inspection
• atomic in-place rewrites
• optional Starlark scripting and a persistent REPL with history, editing, and IDE-style completion for heavier transforms

Install

• Release artifacts: tagged releases publish Linux musl x86_64 and aarch64, macOS aarch64, and Windows x86_64 binaries.
• Nix: nix-build -A bin builds a static Linux binary into ./result/bin/aq.
• Cargo: cargo install --git https://github.com/jpetrucciani/aq

Quick Start

aq '.users[] | select(.age >= 30) | .name' users.json
aq '.name' -
aq -p yaml -o json '.services[].port' compose.yaml
aq -o table '.items[] | {name, status, owner}' data.json
aq --line-stream 'select(.status >= 500)' logs.jsonl
aq --in-place '.version = "2.0"' config.toml
aq -n --starlark '1 + 2'
aq -n -P

Merging files

Use --slurp to read multiple input documents into one array, then merge the first document with the second:

# Shallow object merge. Top-level keys from overlay.yaml replace base.yaml.
aq --slurp '.[0] + .[1]' base.yaml overlay.yaml

# Recursive object merge. Nested objects are merged, deepest values still let
# overlay.yaml win.
aq --slurp '.[0] * .[1]' base.yaml overlay.yaml

# Merge every slurped document recursively, in file order.
aq --slurp 'merge_all(true)' base.yaml env.yaml local.yaml

Docs

• Site source: docs/
• CLI reference: docs/cli.md
• Performance: docs/performance.md
• jq compatibility: docs/jq-compatibility.md
• Starlark: docs/starlark.md
• Starlark examples: examples/starlark/README.md

Development

cargo fmt
cargo clippy --all --benches --tests --examples --all-features -- -D warnings

cd docs
bun install
bun run docs:dev