vm run: ship tracked files only by default; add --include-untracked + --dry-run

Workspace-mode vm run and vm workspace prepare used to copy both
tracked AND untracked non-ignored files into the guest. That silently
catches local .env files, scratch notes, credentials, and any other
working-tree state a developer hasn't explicitly gitignored — a real
data-exposure footgun given the golden image ships Docker and the
usual dev tooling.

Flip the default to tracked-only. Users who actually want the fuller
set opt in with --include-untracked (documented in both commands'
help). Gitignored files are still always excluded regardless of the
flag.

Add --dry-run to both vm run and vm workspace prepare. Dry-run
inspects the repo CLI-side (no VM created, no daemon RPC needed since
the daemon is always local and the inspection is a pure git read),
prints the exact file list + mode, and exits. A byte-level preview of
what would land in the guest.

When running real (non-dry) and untracked files exist in the repo but
are being skipped under the new default, print a one-line notice
pointing to --include-untracked so users aren't surprised when the
guest is missing something they expected.

Signature changes:
- ListOverlayPaths takes an includeUntracked bool (tracked always;
  untracked gated by flag).
- InspectRepo takes the same flag and passes it through.
- VMWorkspacePrepareParams gains IncludeUntracked.
- WorkspaceService.workspaceInspectRepo seam signature widened to
  match (4 callers in tests updated).

New workspace package tests cover both modes and verify that
gitignored files never leak regardless of the flag.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
This commit is contained in:
Thales Maciel 2026-04-21 19:53:17 -03:00
parent 25a1466947
commit 2a7f55f028
No known key found for this signature in database
GPG key ID: 33112E6833C34679
11 changed files with 293 additions and 67 deletions

View file

@ -67,11 +67,12 @@ var HostCommandOutputFunc = func(ctx context.Context, name string, args ...strin
return output, fmt.Errorf("%s: %w: %s", command, err, detail)
}
// InspectRepo resolves rawPath into an absolute repo root and captures the
// HEAD, branch, optional base-from ref, git identity, origin URL, submodules,
// and overlay paths (tracked + untracked non-ignored files) needed for a
// prepare.
func InspectRepo(ctx context.Context, rawPath, branchName, fromRef string) (RepoSpec, error) {
// InspectRepo resolves rawPath into an absolute repo root and captures
// the HEAD, branch, optional base-from ref, git identity, origin URL,
// submodules, and overlay paths needed for a prepare. Overlay paths
// cover tracked files by default; untracked non-ignored files are
// included only when includeUntracked is true.
func InspectRepo(ctx context.Context, rawPath, branchName, fromRef string, includeUntracked bool) (RepoSpec, error) {
sourcePath, err := ResolveSourcePath(rawPath)
if err != nil {
return RepoSpec{}, err
@ -119,7 +120,7 @@ func InspectRepo(ctx context.Context, rawPath, branchName, fromRef string) (Repo
if err != nil {
return RepoSpec{}, fmt.Errorf("resolve origin url for %s: %w", repoRoot, err)
}
overlayPaths, err := ListOverlayPaths(ctx, repoRoot)
overlayPaths, err := ListOverlayPaths(ctx, repoRoot, includeUntracked)
if err != nil {
return RepoSpec{}, err
}
@ -292,17 +293,22 @@ func ListSubmodules(ctx context.Context, repoRoot string) ([]string, error) {
return submodules, nil
}
// ListOverlayPaths returns tracked + untracked non-ignored files in
// repoRoot. Missing tracked entries (deleted working-tree files) are skipped.
func ListOverlayPaths(ctx context.Context, repoRoot string) ([]string, error) {
// ListOverlayPaths returns tracked files in repoRoot, plus (when
// includeUntracked is true) untracked non-ignored files. Missing
// tracked entries (deleted working-tree files) are skipped in both
// modes.
//
// The default is tracked-only because "untracked + not gitignored"
// silently catches local credentials, .env files, scratch notes, and
// other secrets that live in the working tree but aren't meant to
// leave the developer's machine. Callers that genuinely want the
// fuller set (scratch repos, vendored binaries the user is iterating
// on) opt in explicitly.
func ListOverlayPaths(ctx context.Context, repoRoot string, includeUntracked bool) ([]string, error) {
trackedOutput, err := GitOutput(ctx, repoRoot, "ls-files", "-z")
if err != nil {
return nil, fmt.Errorf("list tracked files for %s: %w", repoRoot, err)
}
untrackedOutput, err := GitOutput(ctx, repoRoot, "ls-files", "--others", "--exclude-standard", "-z")
if err != nil {
return nil, fmt.Errorf("list untracked files for %s: %w", repoRoot, err)
}
paths := make([]string, 0)
seen := make(map[string]struct{})
for _, relPath := range ParseNullSeparatedOutput(trackedOutput) {
@ -318,20 +324,44 @@ func ListOverlayPaths(ctx context.Context, repoRoot string) ([]string, error) {
seen[relPath] = struct{}{}
paths = append(paths, relPath)
}
for _, relPath := range ParseNullSeparatedOutput(untrackedOutput) {
if relPath == "" {
continue
if includeUntracked {
untrackedOutput, err := GitOutput(ctx, repoRoot, "ls-files", "--others", "--exclude-standard", "-z")
if err != nil {
return nil, fmt.Errorf("list untracked files for %s: %w", repoRoot, err)
}
if _, ok := seen[relPath]; ok {
continue
for _, relPath := range ParseNullSeparatedOutput(untrackedOutput) {
if relPath == "" {
continue
}
if _, ok := seen[relPath]; ok {
continue
}
seen[relPath] = struct{}{}
paths = append(paths, relPath)
}
seen[relPath] = struct{}{}
paths = append(paths, relPath)
}
sort.Strings(paths)
return paths, nil
}
// CountUntrackedPaths returns the number of untracked non-ignored
// files in repoRoot. Used by the CLI to warn the user when they are
// about to ship a workspace that has local-but-unignored scratch
// files which, under the default, will be skipped.
func CountUntrackedPaths(ctx context.Context, repoRoot string) (int, error) {
untrackedOutput, err := GitOutput(ctx, repoRoot, "ls-files", "--others", "--exclude-standard", "-z")
if err != nil {
return 0, fmt.Errorf("list untracked files for %s: %w", repoRoot, err)
}
count := 0
for _, relPath := range ParseNullSeparatedOutput(untrackedOutput) {
if relPath != "" {
count++
}
}
return count, nil
}
// ParsePrepareMode validates and canonicalises a user-supplied mode value.
func ParsePrepareMode(raw string) (model.WorkspacePrepareMode, error) {
switch strings.TrimSpace(raw) {

View file

@ -0,0 +1,99 @@
package workspace
import (
"context"
"os"
"os/exec"
"path/filepath"
"slices"
"testing"
)
// seedRepo creates a tiny git repo with one tracked file, one
// gitignored file, and one untracked-non-ignored file. Returns the
// repo root path. Skips the test if git isn't on PATH (unusual for
// a dev machine, but polite).
func seedRepo(t *testing.T) string {
t.Helper()
if _, err := exec.LookPath("git"); err != nil {
t.Skipf("git not on PATH: %v", err)
}
dir := t.TempDir()
run := func(args ...string) {
t.Helper()
cmd := exec.Command(args[0], args[1:]...)
cmd.Dir = dir
// Isolate from the ambient user config so commits don't need
// a global user.name/user.email. Also disable GPG signing.
cmd.Env = append(os.Environ(),
"GIT_AUTHOR_NAME=t", "GIT_AUTHOR_EMAIL=t@t",
"GIT_COMMITTER_NAME=t", "GIT_COMMITTER_EMAIL=t@t",
"GIT_CONFIG_GLOBAL=/dev/null",
)
if out, err := cmd.CombinedOutput(); err != nil {
t.Fatalf("%v: %v\n%s", args, err, out)
}
}
writeFile := func(relPath, content string) {
t.Helper()
if err := os.WriteFile(filepath.Join(dir, relPath), []byte(content), 0o644); err != nil {
t.Fatal(err)
}
}
run("git", "init", "-q", "-b", "main")
run("git", "config", "commit.gpgsign", "false")
writeFile(".gitignore", "ignored.log\n")
writeFile("README.md", "hello\n")
run("git", "add", ".gitignore", "README.md")
run("git", "commit", "-q", "-m", "init")
// A tracked file AFTER the first commit so ls-files picks it up.
// A gitignored file so --exclude-standard filters it.
// An untracked non-ignored file so the flag matters.
writeFile("src.go", "package main\n")
run("git", "add", "src.go")
run("git", "commit", "-q", "-m", "src")
writeFile("ignored.log", "noisy\n")
writeFile("SECRETS.env", "TOKEN=abc\n")
return dir
}
func TestListOverlayPaths_TrackedOnlyByDefault(t *testing.T) {
repo := seedRepo(t)
got, err := ListOverlayPaths(context.Background(), repo, false)
if err != nil {
t.Fatalf("ListOverlayPaths: %v", err)
}
want := []string{".gitignore", "README.md", "src.go"}
if !slices.Equal(got, want) {
t.Fatalf("got %v, want %v (untracked SECRETS.env must be excluded; gitignored ignored.log must always be excluded)", got, want)
}
}
func TestListOverlayPaths_IncludeUntracked(t *testing.T) {
repo := seedRepo(t)
got, err := ListOverlayPaths(context.Background(), repo, true)
if err != nil {
t.Fatalf("ListOverlayPaths: %v", err)
}
want := []string{".gitignore", "README.md", "SECRETS.env", "src.go"}
if !slices.Equal(got, want) {
t.Fatalf("got %v, want %v", got, want)
}
// gitignored files must stay out even when untracked is included.
for _, p := range got {
if p == "ignored.log" {
t.Fatalf("gitignored file leaked into overlay: %v", got)
}
}
}
func TestCountUntrackedPaths(t *testing.T) {
repo := seedRepo(t)
count, err := CountUntrackedPaths(context.Background(), repo)
if err != nil {
t.Fatalf("CountUntrackedPaths: %v", err)
}
if count != 1 {
t.Fatalf("count = %d, want 1 (only SECRETS.env; ignored.log is gitignored)", count)
}
}