Merge pull request #533 from AutoMaker-Org/feature/v0.12.0rc-1768605477061-fhv5

fix: Codex freezes

apps/server/src/providers/cli-provider.ts

@@ -35,6 +35,7 @@ import {
   type SubprocessOptions,
   type WslCliResult,
 } from '@automaker/platform';
+import { calculateReasoningTimeout } from '@automaker/types';
 import { createLogger, isAbortError } from '@automaker/utils';
 import { execSync } from 'child_process';
 import * as fs from 'fs';
@@ -107,6 +108,15 @@ export interface CliDetectionResult {
 // Create logger for CLI operations
 const cliLogger = createLogger('CliProvider');
 
+/**
+ * Base timeout for CLI operations in milliseconds.
+ * CLI tools have longer startup and processing times compared to direct API calls,
+ * so we use a higher base timeout (120s) than the default provider timeout (30s).
+ * This is multiplied by reasoning effort multipliers when applicable.
+ * @see calculateReasoningTimeout from @automaker/types
+ */
+const CLI_BASE_TIMEOUT_MS = 120000;
+
 /**
  * Abstract base class for CLI-based providers
  *
@@ -450,6 +460,10 @@ export abstract class CliProvider extends BaseProvider {
       }
     }
 
+    // Calculate dynamic timeout based on reasoning effort.
+    // This addresses GitHub issue #530 where reasoning models with 'xhigh' effort would timeout.
+    const timeout = calculateReasoningTimeout(options.reasoningEffort, CLI_BASE_TIMEOUT_MS);
+
     // WSL strategy
     if (this.useWsl && this.wslCliPath) {
       const wslCwd = windowsToWslPath(cwd);
@@ -473,7 +487,7 @@ export abstract class CliProvider extends BaseProvider {
         cwd, // Windows cwd for spawn
         env: filteredEnv,
         abortController: options.abortController,
-        timeout: 120000, // CLI operations may take longer
+        timeout,
       };
     }
 
@@ -488,7 +502,7 @@ export abstract class CliProvider extends BaseProvider {
         cwd,
         env: filteredEnv,
         abortController: options.abortController,
-        timeout: 120000,
+        timeout,
       };
     }
 
@@ -501,7 +515,7 @@ export abstract class CliProvider extends BaseProvider {
       cwd,
       env: filteredEnv,
       abortController: options.abortController,
-      timeout: 120000,
+      timeout,
     };
   }
 

apps/server/src/providers/codex-provider.ts

@@ -33,6 +33,8 @@ import {
   CODEX_MODEL_MAP,
   supportsReasoningEffort,
   validateBareModelId,
+  calculateReasoningTimeout,
+  DEFAULT_TIMEOUT_MS,
   type CodexApprovalPolicy,
   type CodexSandboxMode,
   type CodexAuthStatus,
@@ -91,7 +93,14 @@ const CODEX_ITEM_TYPES = {
 const SYSTEM_PROMPT_LABEL = 'System instructions';
 const HISTORY_HEADER = 'Current request:\n';
 const TEXT_ENCODING = 'utf-8';
-const DEFAULT_TIMEOUT_MS = 30000;
+/**
+ * Default timeout for Codex CLI operations in milliseconds.
+ * This is the "no output" timeout - if the CLI doesn't produce any JSONL output
+ * for this duration, the process is killed. For reasoning models with high
+ * reasoning effort, this timeout is dynamically extended via calculateReasoningTimeout().
+ * @see calculateReasoningTimeout from @automaker/types
+ */
+const CODEX_CLI_TIMEOUT_MS = DEFAULT_TIMEOUT_MS;
 const CONTEXT_WINDOW_256K = 256000;
 const MAX_OUTPUT_32K = 32000;
 const MAX_OUTPUT_16K = 16000;
@@ -814,13 +823,19 @@ export class CodexProvider extends BaseProvider {
         envOverrides[OPENAI_API_KEY_ENV] = executionPlan.openAiApiKey;
       }
 
+      // Calculate dynamic timeout based on reasoning effort.
+      // Higher reasoning effort (e.g., 'xhigh' for "xtra thinking" mode) requires more time
+      // for the model to generate reasoning tokens before producing output.
+      // This fixes GitHub issue #530 where features would get stuck with reasoning models.
+      const timeout = calculateReasoningTimeout(options.reasoningEffort, CODEX_CLI_TIMEOUT_MS);
+
       const stream = spawnJSONLProcess({
         command: commandPath,
         args,
         cwd: options.cwd,
         env: envOverrides,
         abortController: options.abortController,
-        timeout: DEFAULT_TIMEOUT_MS,
+        timeout,
         stdinData: promptText, // Pass prompt via stdin
       });
 

apps/server/tests/unit/providers/codex-provider.test.ts

@@ -11,6 +11,11 @@ import {
   getCodexConfigDir,
   getCodexAuthIndicators,
 } from '@automaker/platform';
+import {
+  calculateReasoningTimeout,
+  REASONING_TIMEOUT_MULTIPLIERS,
+  DEFAULT_TIMEOUT_MS,
+} from '@automaker/types';
 
 const OPENAI_API_KEY_ENV = 'OPENAI_API_KEY';
 const originalOpenAIKey = process.env[OPENAI_API_KEY_ENV];
@@ -289,5 +294,121 @@ describe('codex-provider.ts', () => {
       expect(codexRunMock).not.toHaveBeenCalled();
       expect(spawnJSONLProcess).toHaveBeenCalled();
     });
+
+    it('passes extended timeout for high reasoning effort', async () => {
+      vi.mocked(spawnJSONLProcess).mockReturnValue((async function* () {})());
+
+      await collectAsyncGenerator(
+        provider.executeQuery({
+          prompt: 'Complex reasoning task',
+          model: 'gpt-5.1-codex-max',
+          cwd: '/tmp',
+          reasoningEffort: 'high',
+        })
+      );
+
+      const call = vi.mocked(spawnJSONLProcess).mock.calls[0][0];
+      // High reasoning effort should have 3x the default timeout (90000ms)
+      expect(call.timeout).toBe(DEFAULT_TIMEOUT_MS * REASONING_TIMEOUT_MULTIPLIERS.high);
+    });
+
+    it('passes extended timeout for xhigh reasoning effort', async () => {
+      vi.mocked(spawnJSONLProcess).mockReturnValue((async function* () {})());
+
+      await collectAsyncGenerator(
+        provider.executeQuery({
+          prompt: 'Very complex reasoning task',
+          model: 'gpt-5.1-codex-max',
+          cwd: '/tmp',
+          reasoningEffort: 'xhigh',
+        })
+      );
+
+      const call = vi.mocked(spawnJSONLProcess).mock.calls[0][0];
+      // xhigh reasoning effort should have 4x the default timeout (120000ms)
+      expect(call.timeout).toBe(DEFAULT_TIMEOUT_MS * REASONING_TIMEOUT_MULTIPLIERS.xhigh);
+    });
+
+    it('uses default timeout when no reasoning effort is specified', async () => {
+      vi.mocked(spawnJSONLProcess).mockReturnValue((async function* () {})());
+
+      await collectAsyncGenerator(
+        provider.executeQuery({
+          prompt: 'Simple task',
+          model: 'gpt-5.2',
+          cwd: '/tmp',
+        })
+      );
+
+      const call = vi.mocked(spawnJSONLProcess).mock.calls[0][0];
+      // No reasoning effort should use the default timeout
+      expect(call.timeout).toBe(DEFAULT_TIMEOUT_MS);
+    });
+  });
+
+  describe('calculateReasoningTimeout', () => {
+    it('returns default timeout when no reasoning effort is specified', () => {
+      expect(calculateReasoningTimeout()).toBe(DEFAULT_TIMEOUT_MS);
+      expect(calculateReasoningTimeout(undefined)).toBe(DEFAULT_TIMEOUT_MS);
+    });
+
+    it('returns default timeout for none reasoning effort', () => {
+      expect(calculateReasoningTimeout('none')).toBe(DEFAULT_TIMEOUT_MS);
+    });
+
+    it('applies correct multiplier for minimal reasoning effort', () => {
+      const expected = Math.round(DEFAULT_TIMEOUT_MS * REASONING_TIMEOUT_MULTIPLIERS.minimal);
+      expect(calculateReasoningTimeout('minimal')).toBe(expected);
+    });
+
+    it('applies correct multiplier for low reasoning effort', () => {
+      const expected = Math.round(DEFAULT_TIMEOUT_MS * REASONING_TIMEOUT_MULTIPLIERS.low);
+      expect(calculateReasoningTimeout('low')).toBe(expected);
+    });
+
+    it('applies correct multiplier for medium reasoning effort', () => {
+      const expected = Math.round(DEFAULT_TIMEOUT_MS * REASONING_TIMEOUT_MULTIPLIERS.medium);
+      expect(calculateReasoningTimeout('medium')).toBe(expected);
+    });
+
+    it('applies correct multiplier for high reasoning effort', () => {
+      const expected = Math.round(DEFAULT_TIMEOUT_MS * REASONING_TIMEOUT_MULTIPLIERS.high);
+      expect(calculateReasoningTimeout('high')).toBe(expected);
+    });
+
+    it('applies correct multiplier for xhigh reasoning effort', () => {
+      const expected = Math.round(DEFAULT_TIMEOUT_MS * REASONING_TIMEOUT_MULTIPLIERS.xhigh);
+      expect(calculateReasoningTimeout('xhigh')).toBe(expected);
+    });
+
+    it('uses custom base timeout when provided', () => {
+      const customBase = 60000;
+      expect(calculateReasoningTimeout('high', customBase)).toBe(
+        Math.round(customBase * REASONING_TIMEOUT_MULTIPLIERS.high)
+      );
+    });
+
+    it('falls back to 1.0 multiplier for invalid reasoning effort', () => {
+      // Test that invalid values fallback gracefully to default multiplier
+      // This tests the defensive ?? 1.0 in calculateReasoningTimeout
+      const invalidEffort = 'invalid_effort' as never;
+      expect(calculateReasoningTimeout(invalidEffort)).toBe(DEFAULT_TIMEOUT_MS);
+    });
+
+    it('produces expected absolute timeout values', () => {
+      // Verify the actual timeout values that will be used:
+      // none: 30000ms (30s)
+      // minimal: 36000ms (36s)
+      // low: 45000ms (45s)
+      // medium: 60000ms (1m)
+      // high: 90000ms (1m 30s)
+      // xhigh: 120000ms (2m)
+      expect(calculateReasoningTimeout('none')).toBe(30000);
+      expect(calculateReasoningTimeout('minimal')).toBe(36000);
+      expect(calculateReasoningTimeout('low')).toBe(45000);
+      expect(calculateReasoningTimeout('medium')).toBe(60000);
+      expect(calculateReasoningTimeout('high')).toBe(90000);
+      expect(calculateReasoningTimeout('xhigh')).toBe(120000);
+    });
   });
 });

libs/types/src/index.ts

@@ -21,6 +21,13 @@ export type {
   ReasoningEffort,
 } from './provider.js';
 
+// Provider constants and utilities
+export {
+  DEFAULT_TIMEOUT_MS,
+  REASONING_TIMEOUT_MULTIPLIERS,
+  calculateReasoningTimeout,
+} from './provider.js';
+
 // Codex CLI types
 export type {
   CodexSandboxMode,

libs/types/src/provider.ts

@@ -18,6 +18,61 @@ import type { CodexSandboxMode, CodexApprovalPolicy } from './codex.js';
  */
 export type ReasoningEffort = 'none' | 'minimal' | 'low' | 'medium' | 'high' | 'xhigh';
 
+/**
+ * Default timeout in milliseconds for provider operations.
+ * Used as the baseline timeout for API calls and CLI operations.
+ */
+export const DEFAULT_TIMEOUT_MS = 30000;
+
+/**
+ * Timeout multipliers for reasoning effort levels.
+ * Higher reasoning effort requires more time for the model to generate reasoning tokens.
+ * These multipliers are applied to DEFAULT_TIMEOUT_MS.
+ */
+export const REASONING_TIMEOUT_MULTIPLIERS: Record<ReasoningEffort, number> = {
+  none: 1.0, // No reasoning, baseline timeout
+  minimal: 1.2, // Very quick reasoning, slight increase
+  low: 1.5, // Quick reasoning, moderate increase
+  medium: 2.0, // Balanced reasoning, double baseline
+  high: 3.0, // Extended reasoning, triple baseline
+  xhigh: 4.0, // Maximum reasoning, quadruple baseline
+};
+
+/**
+ * Calculate timeout for provider operations based on reasoning effort.
+ * Higher reasoning effort requires more time for the model to generate reasoning tokens.
+ *
+ * This function addresses GitHub issue #530 where Codex CLI with GPT-5.2 "xtra thinking"
+ * (xhigh reasoning effort) mode would get stuck because the 30-second "no output" timeout
+ * would trigger during extended reasoning phases.
+ *
+ * @param reasoningEffort - The reasoning effort level, defaults to 'none' if undefined.
+ *                          If an invalid value is provided, falls back to multiplier 1.0.
+ * @param baseTimeoutMs - Optional custom base timeout, defaults to DEFAULT_TIMEOUT_MS (30000ms)
+ * @returns The calculated timeout in milliseconds, rounded to the nearest integer
+ *
+ * @example
+ * // Using default base timeout (30000ms)
+ * calculateReasoningTimeout('high') // Returns 90000 (30000 * 3.0)
+ *
+ * @example
+ * // Using custom base timeout
+ * calculateReasoningTimeout('medium', 60000) // Returns 120000 (60000 * 2.0)
+ *
+ * @example
+ * // No reasoning effort (default)
+ * calculateReasoningTimeout() // Returns 30000 (default timeout)
+ * calculateReasoningTimeout(undefined) // Returns 30000
+ */
+export function calculateReasoningTimeout(
+  reasoningEffort?: ReasoningEffort,
+  baseTimeoutMs: number = DEFAULT_TIMEOUT_MS
+): number {
+  const effort = reasoningEffort ?? 'none';
+  const multiplier = REASONING_TIMEOUT_MULTIPLIERS[effort] ?? 1.0;
+  return Math.round(baseTimeoutMs * multiplier);
+}
+
 /**
  * Configuration for a provider instance
  */