Skip to main content
Versioned File Systems are Git repositories that mount as ordinary directories. Agents can read and write files in a sandbox, snapshot their work, and publish finished changes back to a branch. The repository stays the source of truth. After promotion, future agents that use the branch and Git users who clone or fetch it get the published version. Use versioned file systems when applications need a fast directory interface, but your product still needs Git history, branches, and reproducible versions. Git Repository Dashboard

Quickstart

Install the Tensorlake CLI and sign in:
curl -fsSL https://tensorlake.ai/install | sh
tl login
Then create a repository, mount it, write a file, snapshot the change, and promote it to main.
1

Create a repository

$ tl git create agent-outputs --default-branch main
created https://git.tensorlake.ai/project_9f3c2a1b/agent-outputs
The repository is an ordinary Git remote. You can use it through tl fs mounts or through plain git.
2

Mount a writable workspace

$ tl fs mount agent-outputs:main /work
Mounted agent-outputs:main at /work (workspace 3f9a2b7e1c4d)
/work is now an ordinary writable directory. Changes stay isolated from main until you promote them.
3

Write and snapshot

$ echo "# Parser Notes" > /work/NOTES.md
$ tl fs snapshot /work -m "add notes"
Snapshot 8b21f6a9c3d5e7f1a2b4c6d8e0f3a5b7c9d1e3f5 (1 file(s), 1 of 1 chunks uploaded)
A snapshot is a commit on the workspace. It is private until promotion.
4

Promote to main

$ tl fs promote /work main -m "add notes"
Promoted workspace 3f9a2b7e1c4d -> main at 1c4d9a2f7e5b3d8c6a4f2e0d9b7c5a3f1e8d6b4c (squashed)
After promotion, main contains the notes. New agents that mount agent-outputs:main, and anyone who clones or fetches the repository, get those files. Agents with following read-only mounts pick up the update automatically. The activity history shows who published it and which workspace it came from.

Mental Model

The workflow is:
  1. A repository stores the durable Git history.
  2. A workspace gives an agent a private writable view of that repository.
  3. A mount exposes the workspace as a directory, such as /work.
  4. A snapshot checkpoints the mounted files.
  5. Promotion publishes a snapshot to a branch so future agents and Git users can use that version.
See Core Concepts for short definitions of each term.

What Tensorlake Provides

  • Plain Git repositories: clone, branch, commit, merge, push, and fetch over Git smart HTTP.
  • Writable workspaces: isolated agent workspaces that survive sandbox restarts and can be reattached.
  • Snapshots: incremental checkpoints that record a workspace’s file state as Git commits.
  • Promotion: publish workspace snapshots as Git commits on shared branches, with conflict handling and activity attribution.
  • Read-only mounts: serve a fixed commit or follow a branch across many running sandboxes.

Use Cases

Distribute Files to Agents

Roll out manuals, skills, configs, and tools to many agents with versioned read-only mounts.

Store Agent-Generated Code

Store generated apps, docs, and assets with snapshots, promotion, and activity history.

Where To Go Next

Core Concepts

Learn the vocabulary: repositories, workspaces, mounts, snapshots, promotion, and mount modes.

File System Mounts

Choose between writable, read-only, pinned, and following mounts.

Writable Workspaces

Mount, snapshot, and promote agent work.

Git Repositories

Use Tensorlake repositories with ordinary Git commands.

Repository SDKs

Create repositories, push worktrees, and merge branches from Python or TypeScript.