Hmbown/CodeWhale · error · anyhow::Error

one or more MCP servers failed validation

Error message

one or more MCP servers failed validation

What it means

The `codewhale mcp validate` subcommand checks all configured MCP servers and prints each named failure to stderr; when any validation error exists it aborts with this anyhow bail instead of continuing. It is a deliberate fail-loud gate: the config is considered unusable until every server definition passes validation.

Solutions

1. Read the per-server lines printed above the bail (`- {name}: {err:#}`) to see exactly which server and why.
2. Fix or remove the offending server entry with `codewhale mcp remove <name>` and re-add it correctly.
3. Re-run `codewhale mcp validate` until it reports no failures before starting the TUI.

Example fix

// before (config.json)
{"servers": {"fs": {"command": "/usr/local/bin/mcp-fs"}}}
// after (binary actually exists)
{"servers": {"fs": {"command": "/home/user/.local/bin/mcp-fs"}}}

let ok = std::process::Command::new("codewhale").args(["mcp", "validate"]).status()?.success();
if !ok { eprintln!("fix MCP config before proceeding"); }

match codewhale_mcp::validate_config(&cfg) {
    Ok(()) => start_tui(),
    Err(e) => eprintln!("MCP validation failed: {e:#}"),
}

When it happens

Trigger: Running `codewhale mcp validate` (or a flow that invokes it) when `errors` collected from validating `cfg.servers` is non-empty — e.g. a server entry has a bad command, missing required fields, or an unresolvable workspace.

Common situations: Hand-editing mcp config JSON with a typo, adding a server via `codewhale mcp add` with an invalid command path, or copying config from another machine where paths/binary names don't exist.

Understand the failure class

Background: Schema validation failed / invalid input schema: payload rejected because its shape doesn't match the expected schema — this error's family across 28 libraries.

Related errors

• Provide either a command or URL for MCP server
• bundle at is bytes; the limit is bytes
• doctor configuration validation failed; details omitted…
• doctor configuration validation failed; see JSON output
• invalid JSON payload at key

AI-assisted analysis of Hmbown/CodeWhale@73e0f67d83 (2026-09-22). Data as JSON: /api/errors/700573ec9191dae3. Report an issue: GitHub.

Appendix: source

Thrown at crates/tui/src/lib.rs:10262

            println!("Disabled MCP server '{name}'");
            Ok(())
        }
        McpCommand::Validate => {
            let mut pool = McpPool::from_config_path_with_workspace_and_plugins(
                &config_path,
                workspace,
                std::sync::Arc::new(plugins.clone()),
            )?;
            let errors = pool.connect_all().await;
            if errors.is_empty() {
                println!("MCP config is valid. All enabled servers connected.");
                return Ok(());
            }
            eprintln!("MCP validation failed:");
            for (name, err) in errors {
                eprintln!("  - {name}: {err:#}");
            }
            bail!("one or more MCP servers failed validation");
        }
        McpCommand::AddSelf { name, workspace } => {
            let exe_path = std::env::current_exe()
                .map_err(|e| anyhow!("Cannot resolve current binary path: {e}"))?;
            let exe_str = exe_path.to_string_lossy().to_string();

            let mut args = Vec::with_capacity(if workspace.is_some() { 4 } else { 2 });
            if let Some(ref ws) = workspace {
                args.push("--workspace".to_string());
                args.push(ws.clone());
            }
            args.push("serve".to_string());
            args.push("--mcp".to_string());

            crate::mcp::mutate_config(&config_path, None, |cfg| {
                if cfg.servers.contains_key(&name) {
                    bail!(
                        "MCP server '{name}' already exists in {}. Use `codewhale mcp remove {name}` first, or choose a different --name.",

View on GitHub (pinned to 73e0f67d83)