What causes "git add failed with {}" in nikivdev/code?

Run `git add -- ` manually to see the underlying git error

git add failed with {}

Error message

git add failed with {}

What it means

Staging files for the doc commit via `git add -- <paths>` failed with a non-zero exit status, which the library converts into this error. The context line "failed to run git add" distinguishes spawn failures from this exit-status failure.

Source

Thrown at src/codex_session_docs.rs:1509

    paths.push(display_repo_relative(
        &project_root.display().to_string(),
        &promotion_path_for_session_json(Path::new(&entry.session_json_path))
            .display()
            .to_string(),
    ));
    dedupe_preserve_order(paths)
}

fn git_add_paths(project_root: &Path, paths: &[String]) -> Result<()> {
    let status = Command::new("git")
        .current_dir(project_root)
        .arg("add")
        .arg("--")
        .args(paths)
        .status()
        .context("failed to run git add")?;
    if !status.success() {
        bail!("git add failed with {}", status);
    }
    Ok(())
}

fn git_commit_paths(project_root: &Path, message: &str) -> Result<()> {
    let status = Command::new("git")
        .current_dir(project_root)
        .args(["commit", "-m", message])
        .status()
        .context("failed to run git commit")?;
    if !status.success() {
        bail!("git commit failed with {}", status);
    }
    Ok(())
}

fn build_doc_commit_message(
    entries: &[DocReviewQueueEntry],

View on GitHub (pinned to a747e741ae)

Solutions

Run `git add -- <paths>` manually to see the underlying git error
Remove a stale `.git/index.lock` if present and retry
Ensure the repository is writable by the user running the tool

Example fix

// before
let status = cmd.status().context("failed to run git add")?;
if !status.success() { bail!("git add failed with {}", status); }
// after
if !status.success() {
    let err = String::from_utf8_lossy(&output.stderr);
    bail!("git add failed with {}: {}", status, err.trim());
}

Defensive patterns

Strategy: retry

Validate before calling

anyhow::ensure!(!Path::new(".git/index.lock").exists(), "git index locked; wait or remove stale lock");
anyhow::ensure!(writable(root), "repo must be writable to stage files");

Type guard

fn can_stage(root: &Path) -> bool {
    root.join(".git").exists() && !root.join(".git/index.lock").exists()
}

Try / catch

for attempt in 0..3 {
    match git_add_paths(root, &paths) {
        Ok(()) => break,
        Err(e) if e.to_string().contains("git add failed") && attempt < 2 => {
            std::thread::sleep(Duration::from_millis(250 * (attempt + 1)));
        }
        Err(e) => return Err(e),
    }
}

Prevention

Avoid running concurrent git operations on the same checkout
Ensure the tool's user owns or can write the repository
Retry transient failures after clearing stale index locks

When it happens

Trigger: `git add` failing while staging the enumerated changed paths — commonly due to pathspec errors, permission problems, index lock, or files changed by another process between status and add.

Common situations: Read-only checkout; concurrent git operations holding `.git/index.lock`; paths containing characters your git version mishandles; detached/bare repo states blocking index writes.

Understand the failure class

Background: "git command failed": what it means when a tool shells out to git and git exits non-zero — this error's family across 21 libraries.

Related errors

AI-assisted analysis of nikivdev/code@a747e741ae (2026-09-01). Data as JSON: /api/errors/8bf57d5964b4d841. Report an issue: GitHub.