fix: adjust confirm test for Windows

.github/workflows/run_tests.yaml

@@ -47,11 +47,15 @@ jobs:
       - name: Install gptscript
         run: |
           curl https://get.gptscript.ai/releases/default_windows_amd64_v1/gptscript.exe -o gptscript.exe
+      - name: Create config file
+        run: |
+          echo '{"credsStore":"file"}' > config
       - name: Install dependencies
         run: npm install
       - name: Run Tests
         env:
           GPTSCRIPT_BIN: .\gptscript.exe
+          GPTSCRIPT_CONFIG_FILE: .\config
           OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
           NODE_GPTSCRIPT_SKIP_INSTALL_BINARY: true
         run: npm test

tests/gptscript.test.ts

@@ -401,21 +401,24 @@ describe("gptscript module", () => {
 	}, 10000)
 
 	test("confirm", async () => {
-		let confirmFound = false
 		const t = {
 			instructions: "List the files in the current working directory.",
 			tools: ["sys.exec"]
 		}
+
+		const commands = [`"ls"`, `"dir"`]
+		let confirmCallCount = 0
 		const run = await g.evaluate(t, {confirm: true})
 		run.on(gptscript.RunEventType.CallConfirm, async (data: gptscript.CallFrame) => {
-			expect(data.input).toContain(`"ls"`)
-			confirmFound = true
+			// On Windows, ls is not always a command. The LLM will try to run dir in this case. Allow both.
+			expect(data.input).toContain(commands[confirmCallCount])
+			confirmCallCount++
 			await g.confirm({id: data.id, accept: true})
 		})
 
 		expect(await run.text()).toContain("README.md")
 		expect(run.err).toEqual("")
-		expect(confirmFound).toBeTruthy()
+		expect(confirmCallCount > 0).toBeTruthy()
 	})
 
 	test("do not confirm", async () => {