What version of Codex CLI is running?
0.142.4
What subscription do you have?
plus
Which model were you using?
No response
What platform is your computer?
No response
What terminal emulator and version are you using (if applicable)?
No response
Codex doctor report
What issue are you seeing?
Environment
- Codex CLI: 0.142.4
- Installation: npm
- Node.js: v22.22.1
- macOS: 14.7.4 Sonoma
- Hardware: Intel MacBook Pro (MacBookPro16,1)
- Architecture: x86_64
Config
config.toml
model = "gpt-5.5"
model_reasoning_effort = "medium"
No custom base_url.
No custom model_provider.
Using official ChatGPT login.
Problem
When using GPT-5.5, Codex consistently crashes after entering the "Working..." stage.
Example:
• Working (4s • esc to interrupt)
zsh: trace trap codex
Changing only the model to GPT-5.4 makes the exact same project and prompt work normally.
Expected behavior
Codex should continue processing the request.
Actual behavior
The Codex process exits with SIGTRAP.
Shell output:
Crash report
The crash report shows:
Exception Type: EXC_BREAKPOINT (SIGTRAP)
CPU Type: X86-64
The stack indicates the crash occurs during V8 initialization:
v8::base::OS::SetPermissions(...)
v8::internal::CodeRange::InitReservation(...)
v8::internal::Heap::SetUp(...)
v8::internal::Isolate::Init(...)
v8::Isolate::Initialize(...)
v8::Isolate::New(...)
This suggests the crash happens during V8 isolate initialization rather than during network communication.
Reproduction
- Install Codex CLI 0.142.4
- Login with ChatGPT account
- Set
- Open any project
- Ask Codex to analyze the project
- Wait until "Working..."
- Codex exits with SIGTRAP
Changing only the model to GPT-5.4 avoids the crash.
Additional observations
- GPT-5.4 works correctly.
- GPT-5.5 crashes consistently.
- Same machine.
- Same project.
- Same configuration except model.
- No custom API endpoint.
What steps can reproduce the bug?
Execute the command
What is the expected behavior?
No response
Additional information
No response
What version of Codex CLI is running?
0.142.4
What subscription do you have?
plus
Which model were you using?
No response
What platform is your computer?
No response
What terminal emulator and version are you using (if applicable)?
No response
Codex doctor report
What issue are you seeing?
Environment
Config
config.toml
No custom base_url.
No custom model_provider.
Using official ChatGPT login.
Problem
When using GPT-5.5, Codex consistently crashes after entering the "Working..." stage.
Example:
Changing only the model to GPT-5.4 makes the exact same project and prompt work normally.
Expected behavior
Codex should continue processing the request.
Actual behavior
The Codex process exits with SIGTRAP.
Shell output:
Crash report
The crash report shows:
The stack indicates the crash occurs during V8 initialization:
This suggests the crash happens during V8 isolate initialization rather than during network communication.
Reproduction
Changing only the model to GPT-5.4 avoids the crash.
Additional observations
What steps can reproduce the bug?
Execute the command
What is the expected behavior?
No response
Additional information
No response