## Observed symptom bug(docker): sandboxes never become Ready on WSL 2 + Docker Desktop (supervisor dials [local]) As a developer on Windows using WSL 2 with Docker Desktop (listed as an experimental platform in the support matrix), I want a
## Observed symptom Docker Desktop 4.93.0 Dashboard fails: GPU child crashes normally; renderer crashes with --disable-gpu-sandbox This appears related to #551, but the environment and failure mode differ. In my case, Docker Desktop 4.93.0 still fails on Windows 11 26H2
## Observed symptom [Bug]: pyproject.name conflict b29030ae2ecbe7dcc19d1eb9b0aa55fe9689b992 The value for `pyproject.name` is `qoffee` which is the same as a [published PyPI package](https://pypi.org/project/Qoffee). I realize it is highly unlikely, but this might be a problem if
## Observed symptom [Bug]: Windows: Process identity PowerShell timeout (3s) and schema compile timeout (250ms) cause proxy crash & HTTP 500 Pre-submission checklist - [x] I have searched existing issues to ensure this bug hasn't been reported before - [x]
## Observed symptom 1.7.0 regression: list_pages fails permanently after selected page is closed Description of the bug Regression in 1.7.0 (noted in #2033, where a new issue was requested). After the selected page is closed by something other than this
## Observed symptom [Bug] nousresearch/hermes-sandbox:desktop declares HOME=[local-path] but never creates the directory — browser tool bootstrap hangs (120s timeout) in every fresh sandbox hermes-sandbox:desktop image declares HOME under /home but never creates that directory, b
## Observed symptom [Bug] Installer test fails to create a venv with uv-managed Python on macOS Developer Experience & Ecosystem `test_dev_install_refreshes_cached_catalog_and_keeps_dependency_cache` fails during virtual environment creation when run with uv-managed CPython 3.12.
## Observed symptom cloudflare_zero_trust_dns_location: Value Conversion Error for unset endpoint networks - [x] This is a bug with an existing resource and is not a feature request or enhancement. Feature requests should be submitted with Cloudflare Support or your account
## Observed symptom MCP OAuth: concurrent processes race refresh-token rotation in .credentials.json; next startup hits invalid_grant and demands re-login What version of Codex CLI is running? `codex-cli 0.157.1` (npm), Linux x86_64, KDE/Wayland. Home on persistent ext4. `mcp_oau
## Observed symptom MCP OAuth: after a 5xx on refresh, every later refresh returns 400 until manual re-login The Cloudflare MCP server (`cloudflare-api`) had been refreshing its token hourly without trouble after a fresh login. Then: - One refresh got
## Observed symptom [Bug]: Linux WebKit screenshot hangs on document.fonts.ready in 1.63.0 Playwright 1.63.0, bundled WebKit 26.6 (revision 2359). **Updated September 30 with a runnable public-page reproduction.** The same failure reproduces on the public, unauthenticated https:/
## Observed symptom getPlatformProxy with remote D1 bindings returns error 1031 on edge-preview exchange What versions & operating system are you using? Environment: Wrangler 4.146.0 Account: rob-powers.workers.dev Worker: ravehacks-preview-bootstrap Date: October 1, 2026, 22:46:
## Observed symptom ChatGPT OAuth Error 400 when signing in to OpenAi when trying to add OpenAI provider I get the following error after authorizing from chatgpt: `Error: Failed to login to OpenAI: OpenAI OAuth token request failed (400): {`
## Observed symptom [Bug] SseServerTransport.close() fails with active sessions and leaks the HTTP server port Description SseServerTransport.close() fails whenever at least one SSE session is open. The HTTP server is never closed, so the port stays bound for the lifetime
## Observed symptom [BUG] Tool Execution Validation Error — Databricks Claude rejects tool_result block missing matching tool_use id **Steps to reproduce:** 1. Configure a pipeline/agent using the **Databricks Claude Haiku 4.5** model with a tool that can return an **empty
## Observed symptom [Bug] Built-in grep_app and websearch remote MCPs fail with SSE error 405 (Non-200 status) on every session On every OpenCode session start, both of OmO's built-in **remote** MCPs fail, and the TUI surfaces an error toast: `grep_app