Digest / November 16, 2025
Codex Issues Digest — November 16, 2025
19 new issues filed ·
34 resolved in openai/codex (UTC day).
New issues (19)
- codex exec resume --last rejects prompt when --json is set
- Harness interrupting shell execution and/or truncating output
- It hits the token limit.
- Misleading “still working” messages after “Finished working” (no background tasks actually run)
- Bug: Missing required parameter errors
- Show Credits on VS Code Extension
- Terminal Buffer Logging
- 10 hours of thinking
- Filtering `codex resume` sessions if inside trusted repository
- stream disconnected before completion: failed to parse ErrorResponse: invalid type: null, expected struct Error
- Codex CLI treating interal tool calls as MCP tool calls - requiring truncation and reading 200 lines maximum in local files
- Notification sound when a local job is finished
- Allow binding to local addresses
- ability to auto accept when implementing a big feature instead of keeping hitting accept.
- Bug: Agent stops working due to limit, even though limit isn't reached.
- codex approvals issue
- GPT-5.1-codex doesn't attempt to debug or theorize what's going wrong at all
- @!filename override to force the reading of an entire file
- Weird special characters in web search and code citations in outputted response
Resolved issues (34)
- Latest version of Codex is taking inordinately long time > 9 hours on code refactoring
- VS Code Extension pre-release broken on windows, infinite loading
- Add clarifai as Model provider
- Codex agent unable to use any built in tools including shell tool
- What a shitty subscription payment system you have got. Absolute worst!
- Codex CLI should run brew update before brew upgrade (Brew cask update fails without sync)
- Codex is charging excessive credits for use, not consistent with 1 week ago
- Codex Hanging while working
- codex-cli 0.55.0 opens in a zombie state
- FEATURE REQUEST: /branch command
- Bug: /status always reflects the newest login instead of the current session
- Please make your model stop being stubborn, lazy and dumb as it's a nightmare working with lately
- Bug: Codex chat breaks when it received a pasted from clipboard image that it cannot render
- Terminal - Windows
- Feature Request: Add optional Instructions to settings or config
- Error reading image causes the whole thread to stop and not be able to continue
- "Reaching Limit Do you want to switch to mini" window hides response to previous question and makes it inaccessible even after it's gone.
- Cannot use GPT-5.1-codex from latest alpha
- 5.1 is not good
- Stuck on Thinking...
- model_reasoning_effort is Not Supported in gpt-5.1-codex
- codex-cli 0.20.0 cannot read back any output of shell execution w/ macOS 26 beta
- TUI shows and enforces login for local LLM providers (no ChatGPT, no API key), blocking usage
- GPT-5 CODEX
- Feature: Allow password-less sudo when running codex in docker container
- feat: yes allowed apps (whitelist) please
- Feature: Docker container persist
- Feature Request: Easy Way to Add/Remove Files for Context And Specify Maximum Context Length
- Explain shell command
- Sync ChatGPT Project Memory with Codex for Seamless Ideation-to-Code Continuity
- Codex CLI treating interal tool calls as MCP tool calls - requiring truncation and reading 200 lines maximum in local files
- Notification sound when a local job is finished
- MCP client for `playwright` failed to start: program not found
- GPT-5.1-codex doesn't attempt to debug or theorize what's going wrong at all