No script code available to test. Please ensure you have a s

Error message

No script code available to test. Please ensure you have a script open in the editor.

What it means

The test-run tool for scripts requires script options (code/path) to be present in the current chat context before it can build and execute a test run. When none are found it throws this message telling the user to open a script in the editor first. It also sets a tool status so the model knows no script was available.

Source

Thrown at frontend/src/lib/components/copilot/chat/script/core.ts:900

				content: 'Failed to apply code changes',
				error: errorMessage
			})
			throw new Error(`Failed to apply code changes: ${errorMessage}`)
		}
	}
}

export const testRunScriptTool: Tool<ScriptChatHelpers> = {
	def: TEST_RUN_SCRIPT_TOOL,
	fn: async function ({ args, workspace, helpers, toolCallbacks, toolId }) {
		const scriptOptions = helpers.getScriptOptions()

		if (!scriptOptions) {
			toolCallbacks.setToolStatus(toolId, {
				content: 'No script available to test',
				error: 'No script found in current context'
			})
			throw new Error(
				'No script code available to test. Please ensure you have a script open in the editor.'
			)
		}

		const parsedArgs = await buildTestRunArgs(args, this.def)

		return executeTestRun({
			jobStarter: () =>
				JobService.runScriptPreview({
					workspace: workspace,
					requestBody: {
						path: scriptOptions.path,
						content: scriptOptions.code,
						args: parsedArgs,
						language: scriptOptions.lang as ScriptLang
					}
				}),
			workspace,

View on GitHub (pinned to e474e8803c)

Solutions

Open a script in the editor so the chat context has script code/path, then ask the model to test again.
Start the copilot chat from within a script editor rather than a global/dashboard page.
If a script should be in context, check that the editor tab is focused/saved so helpers can resolve its path and hash.
For flow-related testing, switch the chat to flow mode where the test-run tool can target flow steps instead.

Example fix

// before
// chat opened with no editor content; tool throws
// after
const code = editor.getValue()
if (code.trim()) {
  await chat.testRun({ path: scriptPath })
} else {
  openScript('scripts/foo.py')
}

Defensive patterns

Strategy: validation

Validate before calling

const scriptOptions = getScriptChatOptions()
if (!scriptOptions?.path || !scriptOptions?.code?.trim()) {
  // do not call the tool; open a script first
  return 'Open a script in the editor before asking the assistant to test it.'
}

Type guard

function hasScriptContext(opts: unknown): opts is { path: string; code: string } {
  return !!opts && typeof opts === 'object' &&
    typeof (opts as any).path === 'string' && (opts as any).path.length > 0 &&
    typeof (opts as any).code === 'string' && (opts as any).code.trim().length > 0
}

Try / catch

try {
  await testRunScriptTool.fn({ args, workspace, helpers, toolCallbacks, toolId })
} catch (e) {
  if (e instanceof Error && e.message.includes('No script code available to test')) {
    // prompt user to open a script, then retry
  }
}

Prevention

Open the target script in the editor before starting the copilot chat.
Verify the chat header shows the script path before asking for test runs.
Don't close the editor tab mid-conversation.
For non-script artifacts, switch the chat to the matching mode (flow/app).

When it happens

Trigger: testRunScriptTool.fn was invoked while scriptOptions was null/undefined — i.e. the chat context contains no script (no open editor tab, or the helpers could not resolve a script path/hash for the current context).

Common situations: User opened the AI chat from a contextless view (e.g. an empty editor tab or a non-script page) and asked the model to 'test it'; the editor tab was closed mid-chat; the chat was started before any script was loaded so helpers had nothing to bind to.

Related errors

AI-assisted analysis of windmill-labs/windmill@e474e8803c (2026-09-03). Data as JSON: /api/errors/c86670cb497050bf. Report an issue: GitHub.