Fix --clone flag not working when starting workspace

[gricha]

Summary

• Fixed the --clone flag not working when starting a workspace with perry start <name> --clone <url>
• The root cause was that the container's init command (which handles git cloning) was never being called during container startup
• Added integration test that verifies the repository is actually cloned inside the workspace

Changes

• perry/internal/src/commands/entrypoint.ts: Call runInit() after Docker daemon is ready
• test/integration/cli.test.ts: Enhanced clone test to wait for init completion and verify the repo exists

Test plan

• bun run validate passes
• Manual test: perry start test-clone --clone https://github.com/octocat/Hello-World creates workspace with cloned repo
• Integration test verifies .workspace-initialized marker exists and Hello-World/.git directory is present

🤖 Generated with Claude Code

[sentry]

Bug: The call to await runInit() in the container entrypoint lacks error handling, which can cause an infinite container restart loop on initialization failure, preventing SSH access.
_{Severity: CRITICAL | Confidence: High}

🔍 Detailed Analysis

The runEntrypoint function calls await runInit() without a try...catch block. If runInit() throws an error, for instance due to a failed git clone or a failing bootstrap script, the exception is not caught locally. This causes the entire container process to exit. Because the container is configured with a restartPolicy: 'unless-stopped', Docker will immediately restart it, leading to an infinite loop of startup, failure, and restart. The SSH daemon is started after runInit(), so this failure loop prevents it from ever running, making the workspace inaccessible for debugging.

💡 Suggested Fix

Wrap the await runInit() call within a try...catch block. In the catch block, log the initialization error but do not exit the process. This will allow the container to finish starting up and enable the SSH daemon, letting users connect to the workspace to diagnose the problem.

🤖 Prompt for AI Agent

Review the code at the location below. A potential bug has been identified by an AI
agent.
Verify if this is a real issue. If it is, propose a fix; if not, explain why it's not
valid.

Location: perry/internal/src/commands/entrypoint.ts#L22-L23

Potential issue: The `runEntrypoint` function calls `await runInit()` without a
`try...catch` block. If `runInit()` throws an error, for instance due to a failed git
clone or a failing bootstrap script, the exception is not caught locally. This causes
the entire container process to exit. Because the container is configured with a
`restartPolicy: 'unless-stopped'`, Docker will immediately restart it, leading to an
infinite loop of startup, failure, and restart. The SSH daemon is started after
`runInit()`, so this failure loop prevents it from ever running, making the workspace
inaccessible for debugging.

_{Did we get this right? 👍 / 👎 to inform future reviews.
Reference ID: 8299098}