Git LFS object recovery

A Git LFS repository can retain valid pointers while the large-file objects they reference are missing or corrupt.

Git LFS stores small pointer files in Git while large content lives in local LFS object storage and configured remotes. A clone can therefore have complete Git history but unavailable assets.

Understand the symptom

The pointer's object ID is the expected content hash for a separate LFS object.

WHY

What is happening

Git LFS fsck checks that objects exist and match expected hashes; corrupt files can be moved into .git/lfs/bad. Recovery requires locating the exact object from an authorized remote, clone, cache, artifact store, backup, or recoverable disk.

What you may see

Symptoms that change the recovery path.

01

Object does not exist on server

The pointer is committed but the content was never uploaded or was removed.

02

Smudge error leaves pointer text

Checkout could not retrieve the LFS object.

03

LFS fsck reports corruption

Local bytes do not match the pointer hash.

04

Only some clones open assets

Those systems may hold unique cached objects.

Protect the source

Actions that can make the situation worse.

01

History rewriting changes object references

Preserve all pointers and commits first.

02

Cache cleanup can delete the last object

Freeze prune, cleanup, and CI rotation.

03

Renaming a file does not fix its hash

Match content to the exact LFS object ID.

Do this now

Safe steps before evaluation.

01

Copy every LFS store

Keep .git/lfs/objects, bad-object folders, clones, caches, artifacts, and backups.

02

Inventory missing object IDs

Record pointer OIDs, sizes, paths, revisions, and fsck output.

03

Preserve configuration

Keep .gitattributes, .lfsconfig, remote URLs, include and exclude rules, and versions.

04

Protect affected storage

Image media before cache scans if it reports corruption or read errors.

NEPFOSYS process

How a data-first evaluation proceeds.

01

Pointer-object inventory

NEPFOSYS maps revisions, pointer OIDs, local stores, remotes, clones, and caches.

02

Hash-verified object recovery

Candidate content is matched to the exact expected object ID and placed into a working duplicate.

03

Repository validation

LFS fsck, checkout, revision coverage, asset sizes, and priority files are checked.

A clear next step

Preserve the original device before another repair or recovery attempt.

Start a case with the exact symptoms, incident timeline, prior attempts, encryption details, and the files that matter most. NEPFOSYS will route the case to the appropriate DFW intake or nationwide mail-in path.

Start with a case review

Questions, answered

What customers ask before starting.

Why do I see a small text file instead of my asset?

It is likely the LFS pointer because the large object was not downloaded.

Can the pointer recreate the missing file?

No. It provides the expected object ID and size, not the original content.

Where might the only object remain?

Check authorized clones, local LFS stores, CI caches, artifacts, backups, and configured remotes before cleanup.

DFW local · Nationwide mail-in

Protect the device. Start with the facts.

Start my recovery case