BoundaryML/baml · error

feedback needs a title; pass --title "..." or a JSON payload

Error message

feedback needs a title; pass --title "..." or a JSON payload with a "title" field

What it means

`baml feedback` requires a non-empty title. If the --json payload's title is missing, non-string, or whitespace-only (and no --title flag supplied a value), the command rejects the payload since an untitled feedback record is useless.

Source

Thrown at baml_language/crates/baml_cli/src/feedback_command.rs:409

                 \"description\" are sent (attach files with --files)",
                unknown.join(", ")
            ));
        }

        // Validate types, not just names: a non-string field would ship to
        // PostHog while the preview and the local record (which read via
        // `as_str`) silently dropped it.
        for key in ["title", "description"] {
            if let Some(value) = obj.get(key)
                && !value.is_string()
            {
                return Err(anyhow::anyhow!("feedback field \"{key}\" must be a string"));
            }
        }

        let title = obj.get("title").and_then(Value::as_str).unwrap_or("");
        if title.trim().is_empty() {
            return Err(anyhow::anyhow!(
                "feedback needs a title; pass --title \"...\" or a JSON payload \
                 with a \"title\" field"
            ));
        }
        Ok(from_json)
    }

    /// Shows exactly what a report contains and how it is attributed.
    fn print_preview(&self, payload: &Value, identified: bool, email: Option<&str>) {
        if identified {
            println!(
                "reporting to Boundary as {}:",
                email.unwrap_or("your account")
            );
        } else {
            println!("reporting to Boundary anonymously:");
        }
        println!();

View on GitHub (pinned to bd85ce9dee)

Solutions

1. Add --title "..." on the command line.
2. Include a non-empty "title" string in the --json payload.
3. If the title comes from a variable, ensure it is non-empty after trimming before invoking the command.

Example fix

# before
baml feedback --json '{"description":"crash on build"}'
# after
baml feedback --title "crash on build" --json '{"description":"stack trace follows"}'

# require a non-empty title before invoking
jq -e '.title != null and (.title | type == "string") and (.title | trim | length > 0)' payload.json \
  && baml feedback --json "$(cat payload.json)" \
  || { echo "title required"; exit 1; }

const hasTitle = (p) => typeof p?.title === 'string' && p.title.trim().length > 0;

try {
  run(`baml feedback --json '${payload}'`);
} catch (e) {
  if (String(e).includes('feedback needs a title')) {
    run(`baml feedback --title "untitled feedback" --json '${payload}'`);
  }
}

When it happens

Trigger: `baml feedback --json '{"description":"x"}'` with no --title, or --json '{"title":" "}' / '{"title":""}' — obj.get("title").as_str().trim() is empty.

Common situations: Passing only a description expecting it to suffice; shell quoting stripping the title; piping JSON built from a template lacking a title field.

Understand the failure class

Background: "missing required argument" and "the following required arguments were not provided": what required-argument errors mean and how to fix them — this error's family across 20 libraries.

Related errors

• unknown feedback field(s) {}; only "title" and "description"
• feedback field "{key}" must be a string
• feedback payload must be a JSON object like {"title": "..."}
• invalid Go SDK import path `{import_path}`
• project name cannot be empty

AI-assisted analysis of BoundaryML/baml@bd85ce9dee (2026-09-12). Data as JSON: /api/errors/de9b10e332cc71cf. Report an issue: GitHub.