nanocoai/nanoclaw · warning

Dropping invalid stored MCP server

Error message

Dropping invalid stored MCP server

What it means

A container_config's stored MCP server entry failed validation and was dropped from the sanitized config. The agent will run without that MCP server; the bad entry remains in the DB until corrected.

Source

Thrown at src/container-config.ts:306

      if (
        server.type !== 'http' &&
        typeof pluginRoot === 'string' &&
        pluginRoot.startsWith(`${CONTAINER_PLUGINS_DIR}/`)
      ) {
        server.pluginRoot = pluginRoot;
      }
      if (server.type !== 'http' && server.cwd && !server.pluginRoot) {
        // cwd resolves against a plugin root; without provenance nothing can
        // resolve it. This strip is the ONLY layer (the runtime passes
        // provenance-less servers through untouched), and the breadcrumb
        // lands in host logs instead of nowhere.
        delete server.cwd;
        log.warn('Stripping cwd from stored MCP server without plugin provenance', { group: groupName, server: name });
      }
      servers[name] = server;
      // eslint-disable-next-line no-catch-all/no-catch-all -- validation failures are data errors, not bugs
    } catch (err) {
      log.warn('Dropping invalid stored MCP server', {
        group: groupName,
        server: name,
        reason: err instanceof Error ? err.message : String(err),
      });
    }
  }
  return servers;
}

/**
 * runtime_tier is an isolation control: dropping an unknown stored value would
 * silently compose the group at the default tier — a weaker boundary than the
 * one the value asked for. Fail closed instead: the group refuses to compose
 * until the stored value is fixed. (A *declared* tier the driver cannot
 * realize is refused separately by validateSpec, against the driver's
 * capabilities.)
 */
function parseRuntimeTier(raw: string | null | undefined, groupName: string): 'container' | 'vm' | undefined {

View on GitHub (pinned to 294ef2aee8)

Solutions

1. Inspect `ncl groups config get --id <group>` and find the server named in the log
2. Remove and re-add the server: `ncl groups config remove-mcp-server` then `add-mcp-server` with current syntax
3. Never edit the DB row directly; use the CLI so validation runs

Example fix

// before
config.mcpServers['foo'] = { cmd: 'npx foo' }  // hand-edited, invalid
// after
ncl groups config add-mcp-server --id <group> --name foo --command npx --args foo

ncl groups config get --id <group>  // review mcpServers entries before upgrade

function looksLikeMcpServer(x: unknown): boolean {
  return !!x && typeof x === 'object' && typeof (x as any).command === 'string';
}

try { sanitize(stored) } catch (e) { /* drop entry, re-add via CLI */ }

When it happens

Trigger: `ncl groups config add-mcp-server` stored an entry whose shape later failed schema validation — invalid command/env shape, missing required fields, or an entry hand-edited in the DB.

Common situations: Version upgrades that tightened MCP server validation; hand-edited container.json or DB rows; entries created by older self-mod tooling.

Related errors

• server name must be 1-64 characters of letters, digits, "_"
• type must be "stdio", "http", or "streamable-http"
• Nothing to update — provide at least one of: --provider, --m
• --name is required
• unsupported transport "sse"

AI-assisted analysis of nanocoai/nanoclaw@294ef2aee8 (2026-08-28). Data as JSON: /api/errors/37a3eb68ff2ff6ea. Report an issue: GitHub.