Budibase/budibase · error · HTTPError

Agent is not properly configured: missing AI config

Error message

Agent is not properly configured: missing AI config

What it means

assertAgentHasValidConfig validates that an agent references an AI config before it can run. If agent.aiconfig is falsy, it throws a 422 HTTPError indicating the agent is not properly configured. This is a pre-flight guard so agents fail fast with a clear message instead of failing deep inside model calls.

Source

Thrown at packages/server/src/sdk/workspace/ai/agents/utils.ts:465

  return {
    successResults,
    successNames,
    semanticFailureNames,
    semanticFailureResults,
  }
}

export function formatIncompleteToolCallError(
  incompleteTools: IncompleteToolCall[]
): string {
  const toolNames = incompleteTools.map(t => t.toolName).join(", ")
  return `The AI model failed to complete tool execution${toolNames ? ` for: ${toolNames}` : ""}. This may be due to a compatibility issue with the selected model. Please try a different model or try again.`
}

export const assertAgentHasValidConfig = async (agent: Agent) => {
  if (!agent.aiconfig) {
    throw new HTTPError(
      "Agent is not properly configured: missing AI config",
      422
    )
  }

  const aiConfig = await sdk.ai.configs.find(agent.aiconfig)
  if (!aiConfig) {
    throw new HTTPError(
      `Agent is not properly configured: AI config "${agent.aiconfig}" not found`,
      422
    )
  }
}

View on GitHub (pinned to a81a902e9a)

Solutions

1. Assign a valid AI config to the agent (set agent.aiconfig to an existing config ID) and save it.
2. In the UI, complete agent setup by choosing a model/AI config before invoking the agent.
3. Catch the 422 HTTPError and prompt the user to configure the agent rather than retrying.

Example fix

// before
await runAgent(agent) // 422: missing AI config
// after
if (!agent.aiconfig) {
  agent.aiconfig = existingConfigId
  await sdk.agents.update(agent)
}
await runAgent(agent)

const isAgentConfigured = (agent: Agent): boolean => !!agent.aiconfig
if (!isAgentConfigured(agent)) {
  throw new HTTPError("Select an AI config for this agent first", 422)
}
await assertAgentHasValidConfig(agent)

const hasAiConfig = (agent: Agent): agent is Agent & { aiconfig: string } =>
  typeof agent.aiconfig === "string" && agent.aiconfig.length > 0

try {
  await assertAgentHasValidConfig(agent)
} catch (err) {
  if (err instanceof HTTPError && err.status === 422 && err.message.includes("missing AI config")) {
    // redirect user to agent configuration screen
  }
}

When it happens

Trigger: Invoking an agent (chat/run flows that call assertAgentHasValidConfig) where the Agent document was saved without selecting an AI config, leaving aiconfig undefined.

Common situations: Agents created before AI config setup was completed; aiconfig field dropped by API clients omitting it; agents imported/copied without their config reference; configs deleted leaving agents pointing at nothing (related but distinct from 467).

Related errors

• LLM not available
• Agent is not properly configured: AI config "${agent.aiconfi
• No license key found
• Workspace DB not found - self-host users using cloud don't h
• CouchDB username not set

AI-assisted analysis of Budibase/budibase@a81a902e9a (2026-08-29). Data as JSON: /api/errors/17d7d54f3dee15ec. Report an issue: GitHub.