git cat-file

The object-database microscope. Learn -p, -t, -s, --batch.

🎯 What & why

git cat-file is the object-database microscope: it reads any object (blob, tree, commit, tag) by SHA and tells you its type, size, or pretty-printed content. It's the first plumbing command everyone should learn because it makes Git's storage stop being magic.

🧠 Mental model

Git is a content-addressed key-value store; cat-file is the GET. Hand it a hash, ask -t (what kind), -s (how big), -p (show me), or -e (does it exist), and it answers from .git/objects.

🛠️ Synopsis

git cat-file (-t | -s | -e | -p) <object>
git cat-file <type> <object>
git cat-file --batch[=<format>] [--follow-symlinks]
git cat-file --batch-check[=<format>]
git cat-file --batch-all-objects [--batch | --batch-check] [--unordered]
git cat-file (--textconv | --filters) [<rev>:<path> | --path=<path> <object>]

🎚️ Switches & options

Flag	What it does
`-t`	Print the object's type (blob, tree, commit, tag).
`-s`	Print the object's size in bytes.
`-p`	Pretty-print the object's contents based on type.
`-e`	Exit 0 if object exists and is valid; silent otherwise. Useful in scripts.
`--batch[=<fmt>]`	Read object names on stdin, write info + content per line. Fast bulk inspection.
`--batch-check[=<fmt>]`	Like --batch but only metadata, no content. Cheap existence/type probes.
`--batch-all-objects`	Iterate every object in the repo (with --batch or --batch-check).
`--filters / --textconv`	Apply smudge filters / textconv for human-readable output of a path.

💡 Use cases

Dump the actual bytes of a blob: git cat-file -p <sha>
Inspect a commit's parents and tree without porcelain noise.
Audit every object in the repo with --batch-all-objects | grep blob.
Verify an object exists in a script with -e (exit code).

🧪 Examples

Pretty-print a commit

git cat-file -p HEAD

Get type of an object

git cat-file -t 4b825dc642cb6eb9a060e54bf8d69288fbee4904

Bulk-check all objects

git cat-file --batch-check --batch-all-objects | head

Read a file at a revision

git cat-file -p HEAD:README.md

🎓 Recommendations

Use -p before reaching for any porcelain when debugging weird state. It never lies.
Reach for --batch-check when scripting; spawning cat-file per object is brutally slow.
Combine with rev-parse (e.g. HEAD^{tree}) to navigate the object graph without guessing.

🪤 Common pitfalls

-p output for trees is human-formatted, not raw bytes; use 'git cat-file tree <sha>' for raw.
Short SHAs can be ambiguous; cat-file will refuse with 'ambiguous'. Pass more characters.
--filters/--textconv require a path context; passing a bare blob SHA won't apply attributes.

🔗 Related modules

📝 Quiz

Hit each option, then Check answers. Score is recorded; Next is always open.