What causes "'codanna init' exited with status {}" in nikivdev/code?

Run `codanna init` manually in the project root and read its stderr output to see the underlying failure.

'codanna init' exited with status {}

Error message

'codanna init' exited with status {}

What it means

ensure_codanna_initialized spawns the external `codanna init` command as a subprocess. If the process exits with a non-zero status (after successfully spawning), the tool wraps that exit code into this anyhow error. It indicates codanna's own initialization step failed, not a spawn failure.

Source

Thrown at src/indexer.rs:64

    let settings = project_root.join(".codanna/settings.toml");
    if settings.exists() {
        return Ok(());
    }

    println!(
        "No Codanna settings found at {} – running 'codanna init'.",
        settings.display()
    );
    let status = Command::new(binary)
        .arg("init")
        .current_dir(project_root)
        .status()
        .with_context(|| "failed to spawn 'codanna init'")?;

    if status.success() {
        Ok(())
    } else {
        bail!(
            "'codanna init' exited with status {}",
            status.code().unwrap_or(-1)
        );
    }
}

fn run_codanna_index(binary: &Path, project_root: &Path) -> Result<()> {
    println!("Indexing project {} via Codanna...", project_root.display());
    let status = Command::new(binary)
        .arg("index")
        .arg("--progress")
        .arg(".")
        .current_dir(project_root)
        .status()
        .with_context(|| "failed to spawn 'codanna index'")?;

    if status.success() {
        Ok(())

View on GitHub (pinned to a747e741ae)

Solutions

Run `codanna init` manually in the project root and read its stderr output to see the underlying failure.
Check `codanna --version` and upgrade/reinstall codanna if it is outdated or broken.
Fix any codanna configuration issues (invalid config file, missing permissions) reported by manual init.
Verify the `codanna` binary resolved on PATH is the intended one (`which codanna`).

Defensive patterns

Strategy: try-catch

Validate before calling

if which::which("codanna").is_err() {
    eprintln!("codanna not found on PATH; install it first");
}
let ok = std::process::Command::new("codanna")
    .arg("--version")
    .status()
    .map(|s| s.success())
    .unwrap_or(false);
if !ok { eprintln!("codanna is broken or too old"); }

Try / catch

match run_indexer() {
    Ok(stats) => handle(stats),
    Err(e) if e.to_string().contains("'codanna init' exited") => {
        eprintln!("codanna init failed; run `codanna init` manually to see stderr");
    }
    Err(e) => return Err(e),
}

Prevention

Install and verify codanna (`codanna --version`) before running the indexer
Run `codanna init` once manually per project and commit the resulting config
Keep codanna up to date to avoid subcommand/behavior drift

When it happens

Trigger: Running the indexer's run() flow when `codanna init` is not properly installed/configured, has a broken config, or otherwise returns a non-zero exit code.

Common situations: codanna is installed but its config file is invalid; a PATH-resolved `codanna` shim is an old incompatible version; codanna init fails due to permissions in the target directory.

Related errors

AI-assisted analysis of nikivdev/code@a747e741ae (2026-09-01). Data as JSON: /api/errors/f6f54a7fa1abd0a6. Report an issue: GitHub.