Fix runClaudeUpdate test failure in claude-runner-integration.test.ts

[frankbria]

Description

The runClaudeUpdate test in tests/e2b/claude-runner-integration.test.ts is failing when E2B_API_KEY is available.

Test Location

• File: tests/e2b/claude-runner-integration.test.ts:60-69
• Test Name: should update Claude to latest version

Error

AssertionError: expected false to be true // Object.is equality

- Expected
+ Received

- true
+ false

 ❯ tests/e2b/claude-runner-integration.test.ts:66:30
     64|       const result = await runClaudeUpdate(sandbox, logger);
     65| 
     66|       expect(result.success).toBe(true);

Root Cause Analysis

The test creates a real E2B sandbox and runs claude update command, expecting it to succeed. The command is returning a non-zero exit code.

Potential causes:

1. Claude CLI not properly installed/available in the E2B sandbox template
2. The claude update command failing due to permission issues (npm global install)
3. Network issues in the sandbox environment
4. The E2B template (anthropic-claude-code) may have changed

Relevant Code

// src/e2b/claude-runner.ts
export async function runClaudeUpdate(
  sandbox: Sandbox,
  logger: Logger,
  authMethod: 'api-key' | 'oauth' = 'api-key'
): Promise<ClaudeUpdateResult> {
  // ... runs 'claude update' command
  const result = await sandbox.commands.run(updateCommand, {
    timeoutMs: CLAUDE_UPDATE_TIMEOUT_MS
  });
  const success = result.exitCode === 0;
  // ...
}

Proposed Fix Options

1. Skip update if already at latest: Check Claude version first, skip update if already current
2. Handle update failure gracefully: Mark test as skipped/warning if update fails due to known E2B limitations
3. Use specific version check: Instead of running update, verify Claude is installed and accessible
4. Update sandbox template expectations: Adjust test expectations based on actual sandbox behavior

Reproduction

E2B_API_KEY=your-key npm test -- tests/e2b/claude-runner-integration.test.ts -t "update Claude" --run

Labels

• bug
• component-e2b
• component-tests
• priority-medium

[coderabbitai]

📝 CodeRabbit Plan Mode

Generate an implementation plan and prompts that you can use with your favorite coding agent.

• Create Plan

Examples

• Example 1
• Example 2

---

⚠️ Possible Duplicate Issue

• [BUG] E2B sandbox requires ANTHROPIC_API_KEY but should use E2B_API_KEY #19

🔗 Related PRs

#7 - Fix sandbox-run command database schema mismatch [merged]
#8 - Fix migration system with automatic updates and improved reliability [merged]
#11 - feat: Use anthropic-claude-code E2B template with fallback installation [merged]
#18 - test: Add comprehensive E2E tests for E2B workflow [merged]

---

🧪 Issue enrichment is currently in open beta.

You can configure auto-planning by selecting labels in the issue_enrichment configuration.

To disable automatic issue enrichment, add the following to your .coderabbit.yaml:

issue_enrichment:
  auto_enrich:
    enabled: false

💬 Have feedback or questions? Drop into our discord!

[traycerai]

Plan

Observations

The runClaudeUpdate test fails because it expects claude update to return exit code 0, but the command may fail for several reasons: (1) Claude CLI is already at the latest version and returns non-zero for "already up-to-date", (2) npm global install permission issues in the E2B sandbox, (3) the command doesn't have the --yes flag that updateClaudeCode uses. The more robust updateClaudeCode function (file:src/e2b/claude-runner.ts lines 393-482) implements multiple fallback strategies, while runClaudeUpdate (lines 885-965) is simpler and less resilient.

Approach

Align runClaudeUpdate with the proven multi-strategy approach used in updateClaudeCode. The fix will make runClaudeUpdate check the current Claude version first, then attempt update with --yes flag, and gracefully handle "already up-to-date" scenarios by treating them as success. This maintains backward compatibility while making the function more resilient to E2B sandbox environment variations. The test will be updated to verify the function handles both successful updates and "already current" scenarios correctly.

Implementation Steps

1. Enhance runClaudeUpdate Function Logic

File: file:src/e2b/claude-runner.ts (lines 885-965)

Modify the runClaudeUpdate function to implement a more robust update strategy:

• Add version check before update: Run claude --version to capture the current version before attempting update
• Try update with --yes flag first: Modify the update command to use claude update --yes to auto-accept prompts, with fallback to claude update if --yes is not supported
• Handle "already up-to-date" scenarios: Parse stdout/stderr for messages indicating Claude is already at the latest version (e.g., "already at latest", "no updates available", "up to date")
• Treat "already current" as success: If the command fails but output indicates Claude is already current, set success = true and use the current version
• Improve version parsing: Extract version from both update output and claude --version command
• Enhanced error logging: Log the full command output for debugging, including both stdout and stderr

The logic flow should be:

1. Check current version with claude --version
2. Attempt claude update --yes (or with ANTHROPIC_API_KEY if api-key mode)
3. If exit code is non-zero, check if output indicates "already up-to-date"
4. If already up-to-date, return success with current version
5. If genuinely failed, return failure with error details

2. Update Test Expectations

File: file:tests/e2b/claude-runner-integration.test.ts (lines 60-69)

Modify the test to handle both update scenarios:

• Accept success for both update and "already current": The test should pass if result.success is true, regardless of whether an actual update occurred
• Add version validation: Verify that result.version is a valid version string (not "unknown")
• Add output validation: Check that result.output contains meaningful information about the update attempt
• Consider adding a second test case: Create a separate test that explicitly checks the "already up-to-date" scenario by running runClaudeUpdate twice in sequence

3. Add Integration Test for "Already Current" Scenario

File: file:tests/e2b/claude-runner-integration.test.ts (after line 69)

Add a new test case to explicitly verify the "already up-to-date" behavior:

• Test name: should handle already up-to-date Claude gracefully
• Test logic:
  1. Run runClaudeUpdate once to ensure Claude is at latest
  2. Run runClaudeUpdate again immediately
  3. Verify second call returns success: true even though no update occurred
  4. Verify version is consistent between both calls
• Assertions: Both calls should succeed, versions should match, output should indicate "already up-to-date" or similar

4. Consider Alternative: Make Test More Lenient

Alternative approach if modifying runClaudeUpdate is not desired:

File: file:tests/e2b/claude-runner-integration.test.ts (lines 60-69)

Modify the test to be more lenient about update failures:

• Change assertion: Instead of expect(result.success).toBe(true), check if either success is true OR the error indicates "already up-to-date"
• Add skip condition: Consider using it.skip() or conditional skip if the E2B template is known to have issues with claude update
• Add warning message: Log a warning if update fails but don't fail the test, since the update step is not critical for testing other functionality

5. Update Documentation

File: file:docs/E2B_GUIDE.md (around line 238)

Update the documentation to clarify the --skip-claude-update option:

• Explain when to use: Document scenarios where skipping update is appropriate (e.g., when Claude is known to be current, or when update fails in specific E2B templates)
• Add troubleshooting section: Include guidance on handling claude update failures
• Reference the test: Mention that the integration test verifies update behavior

Sequence Diagram

sequenceDiagram
    participant Test as Integration Test
    participant RCU as runClaudeUpdate
    participant Sandbox as E2B Sandbox
    
    Test->>RCU: Call runClaudeUpdate(sandbox, logger)
    RCU->>Sandbox: Run 'claude --version'
    Sandbox-->>RCU: Current version (e.g., "1.2.3")
    
    alt API Key Mode
        RCU->>Sandbox: Run 'ANTHROPIC_API_KEY=xxx claude update --yes'
    else OAuth Mode
        RCU->>Sandbox: Run 'claude update --yes'
    end
    
    alt Update Successful
        Sandbox-->>RCU: Exit code 0, "Updated to 1.2.4"
        RCU-->>Test: {success: true, version: "1.2.4"}
    else Already Up-to-Date
        Sandbox-->>RCU: Exit code 1, "Already at latest version"
        RCU->>RCU: Parse output, detect "already" message
        RCU-->>Test: {success: true, version: "1.2.3"}
    else Genuine Failure
        Sandbox-->>RCU: Exit code 1, "Permission denied"
        RCU-->>Test: {success: false, error: "Update failed"}
    end
    
    Test->>Test: Assert result.success === true
    Test->>Test: Assert result.version !== "unknown"

Loading

Files to Modify

1. file:src/e2b/claude-runner.ts - Enhance runClaudeUpdate function (lines 885-965)
2. file:tests/e2b/claude-runner-integration.test.ts - Update test expectations and add new test case (lines 60-69 and after)
3. file:docs/E2B_GUIDE.md - Update documentation about --skip-claude-update option (around line 238)

Success Criteria

• The runClaudeUpdate test passes consistently when E2B_API_KEY is available
• The function handles "already up-to-date" scenarios gracefully
• The function maintains backward compatibility with existing callers
• Test coverage includes both successful update and "already current" scenarios
• Documentation clearly explains when update failures are expected and acceptable

Import In IDE

🤖 Prompt for AI Agents

## Observations

The `runClaudeUpdate` test fails because it expects `claude update` to return exit code 0, but the command may fail for several reasons: (1) Claude CLI is already at the latest version and returns non-zero for "already up-to-date", (2) npm global install permission issues in the E2B sandbox, (3) the command doesn't have the `--yes` flag that `updateClaudeCode` uses. The more robust `updateClaudeCode` function (file:src/e2b/claude-runner.ts lines 393-482) implements multiple fallback strategies, while `runClaudeUpdate` (lines 885-965) is simpler and less resilient.

## Approach

Align `runClaudeUpdate` with the proven multi-strategy approach used in `updateClaudeCode`. The fix will make `runClaudeUpdate` check the current Claude version first, then attempt update with `--yes` flag, and gracefully handle "already up-to-date" scenarios by treating them as success. This maintains backward compatibility while making the function more resilient to E2B sandbox environment variations. The test will be updated to verify the function handles both successful updates and "already current" scenarios correctly.

## Implementation Steps

### 1. Enhance `runClaudeUpdate` Function Logic

**File:** `file:src/e2b/claude-runner.ts` (lines 885-965)

Modify the `runClaudeUpdate` function to implement a more robust update strategy:

- **Add version check before update**: Run `claude --version` to capture the current version before attempting update
- **Try update with `--yes` flag first**: Modify the update command to use `claude update --yes` to auto-accept prompts, with fallback to `claude update` if `--yes` is not supported
- **Handle "already up-to-date" scenarios**: Parse stdout/stderr for messages indicating Claude is already at the latest version (e.g., "already at latest", "no updates available", "up to date")
- **Treat "already current" as success**: If the command fails but output indicates Claude is already current, set `success = true` and use the current version
- **Improve version parsing**: Extract version from both update output and `claude --version` command
- **Enhanced error logging**: Log the full command output for debugging, including both stdout and stderr

The logic flow should be:
1. Check current version with `claude --version`
2. Attempt `claude update --yes` (or with ANTHROPIC_API_KEY if api-key mode)
3. If exit code is non-zero, check if output indicates "already up-to-date"
4. If already up-to-date, return success with current version
5. If genuinely failed, return failure with error details

### 2. Update Test Expectations

**File:** `file:tests/e2b/claude-runner-integration.test.ts` (lines 60-69)

Modify the test to handle both update scenarios:

- **Accept success for both update and "already current"**: The test should pass if `result.success` is true, regardless of whether an actual update occurred
- **Add version validation**: Verify that `result.version` is a valid version string (not "unknown")
- **Add output validation**: Check that `result.output` contains meaningful information about the update attempt
- **Consider adding a second test case**: Create a separate test that explicitly checks the "already up-to-date" scenario by running `runClaudeUpdate` twice in sequence

### 3. Add Integration Test for "Already Current" Scenario

**File:** `file:tests/e2b/claude-runner-integration.test.ts` (after line 69)

Add a new test case to explicitly verify the "already up-to-date" behavior:

- **Test name**: `should handle already up-to-date Claude gracefully`
- **Test logic**: 
  1. Run `runClaudeUpdate` once to ensure Claude is at latest
  2. Run `runClaudeUpdate` again immediately
  3. Verify second call returns `success: true` even though no update occurred
  4. Verify version is consistent between both calls
- **Assertions**: Both calls should succeed, versions should match, output should indicate "already up-to-date" or similar

### 4. Consider Alternative: Make Test More Lenient

**Alternative approach if modifying `runClaudeUpdate` is not desired:**

**File:** `file:tests/e2b/claude-runner-integration.test.ts` (lines 60-69)

Modify the test to be more lenient about update failures:

- **Change assertion**: Instead of `expect(result.success).toBe(true)`, check if either success is true OR the error indicates "already up-to-date"
- **Add skip condition**: Consider using `it.skip()` or conditional skip if the E2B template is known to have issues with `claude update`
- **Add warning message**: Log a warning if update fails but don't fail the test, since the update step is not critical for testing other functionality

### 5. Update Documentation

**File:** `file:docs/E2B_GUIDE.md` (around line 238)

Update the documentation to clarify the `--skip-claude-update` option:

- **Explain when to use**: Document scenarios where skipping update is appropriate (e.g., when Claude is known to be current, or when update fails in specific E2B templates)
- **Add troubleshooting section**: Include guidance on handling `claude update` failures
- **Reference the test**: Mention that the integration test verifies update behavior

## Sequence Diagram

```mermaid
sequenceDiagram
    participant Test as Integration Test
    participant RCU as runClaudeUpdate
    participant Sandbox as E2B Sandbox
    
    Test->>RCU: Call runClaudeUpdate(sandbox, logger)
    RCU->>Sandbox: Run 'claude --version'
    Sandbox-->>RCU: Current version (e.g., "1.2.3")
    
    alt API Key Mode
        RCU->>Sandbox: Run 'ANTHROPIC_API_KEY=xxx claude update --yes'
    else OAuth Mode
        RCU->>Sandbox: Run 'claude update --yes'
    end
    
    alt Update Successful
        Sandbox-->>RCU: Exit code 0, "Updated to 1.2.4"
        RCU-->>Test: {success: true, version: "1.2.4"}
    else Already Up-to-Date
        Sandbox-->>RCU: Exit code 1, "Already at latest version"
        RCU->>RCU: Parse output, detect "already" message
        RCU-->>Test: {success: true, version: "1.2.3"}
    else Genuine Failure
        Sandbox-->>RCU: Exit code 1, "Permission denied"
        RCU-->>Test: {success: false, error: "Update failed"}
    end
    
    Test->>Test: Assert result.success === true
    Test->>Test: Assert result.version !== "unknown"
```

## Files to Modify

1. **`file:src/e2b/claude-runner.ts`** - Enhance `runClaudeUpdate` function (lines 885-965)
2. **`file:tests/e2b/claude-runner-integration.test.ts`** - Update test expectations and add new test case (lines 60-69 and after)
3. **`file:docs/E2B_GUIDE.md`** - Update documentation about `--skip-claude-update` option (around line 238)

## Success Criteria

- The `runClaudeUpdate` test passes consistently when E2B_API_KEY is available
- The function handles "already up-to-date" scenarios gracefully
- The function maintains backward compatibility with existing callers
- Test coverage includes both successful update and "already current" scenarios
- Documentation clearly explains when update failures are expected and acceptable

---

Execution Information

Branch: main
Commit: 14bea90

---

💡 Tips

Supported Commands (Inside Comments)

• Use @traycerai generate to iterate on the previous version of the implementation plan.

Supported Commands (Inside Description)

• Add @traycerai ignore anywhere in the ticket description to prevent this ticket from being processed.
• Add @traycerai branch:<branch-name> anywhere in the ticket description to specify the target branch for the implementation plan.

Community

• Join our Discord Community to get help, request features, and share feedback.
• Follow us on X/Twitter for updates and announcements.