multica-ai/multica · error

failed to save config: %w

Error message

failed to save config: %w

What it means

After a successful browser login the CLI writes the updated profile config (new PAT token, server/app URLs, workspace reset) via SaveCLIConfigForProfile. This error wraps any failure persisting that file, meaning login succeeded server-side but the credential was not saved locally.

Source

Thrown at server/cmd/multica/cmd_auth.go:356

	patClient := cli.NewAPIClient(serverURL, "", patResp.Token)
	var me struct {
		Name  string `json:"name"`
		Email string `json:"email"`
	}
	if err := patClient.GetJSON(ctx, "/api/me", &me); err != nil {
		return cli.WithUserMessage("Sign-in did not complete: the server did not accept the new credential. Run `multica login` again.", err)
	}

	// Save to config. Reset workspace data on every login — the user or
	// server may have changed, so stale workspaces must not persist.
	profile := resolveProfile(cmd)
	cfg, _ := cli.LoadCLIConfigForProfile(profile)
	cfg.WorkspaceID = ""
	cfg.Token = patResp.Token
	cfg.ServerURL = serverURL
	cfg.AppURL = appURL
	if err := cli.SaveCLIConfigForProfile(cfg, profile); err != nil {
		return fmt.Errorf("failed to save config: %w", err)
	}

	fmt.Fprintf(os.Stderr, "Authenticated as %s (%s)\nToken saved to config.\n", me.Name, me.Email)
	return nil
}

func runningInSSHSession() bool {
	for _, key := range []string{"SSH_CONNECTION", "SSH_CLIENT", "SSH_TTY"} {
		if strings.TrimSpace(os.Getenv(key)) != "" {
			return true
		}
	}
	return false
}

func callbackHostFlagValue(cmd *cobra.Command) string {
	for c := cmd; c != nil; c = c.Parent() {
		if value := nonEmptyFlagValue(c.Flags(), callbackHostFlag); value != "" {

View on GitHub (pinned to 2c0912b6ec)

Solutions

1. Check and fix permissions on the profile config directory (e.g. ~/.config/multica or the XDG path) — ensure it is writable by the current user
2. Free disk space if the volume is full
3. If a file exists where the config directory should be, remove/rename it
4. After fixing, rerun `multica login` (the server-side PAT may already exist; re-login is harmless)

Example fix

# diagnose and fix
ls -la ~/.config/multica 2>/dev/null || echo 'missing'
mkdir -p ~/.config/multica && chmod u+w ~/.config/multica
multica login

# verify the config dir is writable before login
cfgdir="${XDG_CONFIG_HOME:-$HOME/.config}/multica"
mkdir -p "$cfgdir" && [ -w "$cfgdir" ] || { echo "config dir not writable: $cfgdir" >&2; exit 1; }

When it happens

Trigger: Config directory not writable (permissions, read-only home); disk full; XDG/config path resolving to a file instead of a directory; SELinux/AppArmor denying writes to the config path.

Common situations: Running the CLI as a different user than the one who owns ~/.config; shared/locked-down machines; containers with read-only home; home quota exceeded.

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
• no input

AI-assisted analysis of multica-ai/multica@2c0912b6ec (2026-08-15). Data as JSON: /api/errors/fc701f56d16ff3b2. Report an issue: GitHub.