multica-ai/multica · error

no input

Error message

no input

What it means

In token login, when --token was not supplied the CLI prompts on stdin with a bufio.Scanner. If scanner.Scan() returns false the input stream is closed (EOF) before any line was read — there is literally no input — so it aborts with 'no input' rather than an empty-token error.

Source

Thrown at server/cmd/multica/cmd_auth.go:425

		fmt.Fprintf(&b, "\nRemote SSH session detected. Before opening that URL on your local computer, forward the callback port in another terminal:\n  ssh -L %d:127.0.0.1:%d <user>@<remote-host>\nThen open the URL above in your local browser.\n", port, port)
	}
	fmt.Fprintln(&b, "\nWaiting for authentication...")
	return b.String()
}

func runAuthLoginToken(cmd *cobra.Command, providedToken string) error {
	// The prompt sentinel is what pflag substitutes for `--token` with no
	// value (see loginCmd init); treat it the same as an empty string so we
	// fall through to the interactive prompt.
	if providedToken == tokenPromptSentinel {
		providedToken = ""
	}
	token := strings.TrimSpace(providedToken)
	if token == "" {
		fmt.Print("Enter your personal access token: ")
		scanner := bufio.NewScanner(os.Stdin)
		if !scanner.Scan() {
			return fmt.Errorf("no input")
		}
		token = strings.TrimSpace(scanner.Text())
	}
	if token == "" {
		return fmt.Errorf("token is required")
	}
	if err := validateLoginTokenPrefix(token); err != nil {
		return err
	}

	serverURL := resolveLoginTokenServerURL(cmd)
	client := cli.NewAPIClient(serverURL, "", token)

	ctx, cancel := cli.APIContext(context.Background())
	defer cancel()

	var me struct {
		Name  string `json:"name"`

View on GitHub (pinned to 2c0912b6ec)

Solutions

1. Always pass the token explicitly in scripts: `multica login --token mul_...`
2. Pipe the token via stdin when prompts are hard: `echo "$TOKEN" | multica login`
3. Run the command in a real terminal when using the interactive prompt

Example fix

# before (non-interactive, stdin closed)
multica login < /dev/null

# after
multica login --token "$MULTICA_TOKEN"

# never rely on the prompt in non-interactive contexts
[ -t 0 ] || : "${MULTICA_TOKEN:?stdin is not a TTY; pass --token}"

When it happens

Trigger: Running `multica login` interactively expecting the token prompt but stdin is closed/redirected from /dev/null; piping nothing (`multica login < /dev/null`); running the command under a non-interactive scheduler with no TTY.

Common situations: Automation invoking the login command without --token; CI jobs; users hitting Enter-less EOF (Ctrl-D) at the prompt; wrappers that detach stdin.

Related errors

• could not start the local login callback server (used to rec
• failed to generate state: %w
• local server error: %w
• timed out waiting for authentication
• failed to save config: %w

AI-assisted analysis of multica-ai/multica@2c0912b6ec (2026-08-15). Data as JSON: /api/errors/ab695f38e324181e. Report an issue: GitHub.