Error listings
Fix OAuth re-authentication failure in GitHub Copilot CLI MCP server when eager-startup OAuth flow conflicts with user-initiated re-auth Includes evidence for GitHub Copilot troubleshooting demand.
Authentication failed: MCPOAuthError: listen EADDRINUSE: address already in use 127.0.0.1:19878 Fix LiteLLM proxy Prisma migration failure when running in non-root Docker container on Alpine/Wolfi-based images Includes evidence for LiteLLM troubleshooting demand.
Prisma migrations not applied with docker non_root image — Prisma warning: doesn't know which engines to download for Linux distro wolfi, falling back to debian builds Fix subagent dispatch pattern where coordinator agent's vault-based OAuth credentials cannot reach subagent's MCP tool execution context Includes evidence for Anthropic API troubleshooting demand.
MCP OAuth calls fail in subagents — vault_ids parameter exists only on sessions.create, no mechanism to pass credentials when dispatching subagents via agent_toolset Fix anthropic SDK client.copy() method that propagates parent api_key into child clients, preventing credential provider (OAuth, GCP, AWS) authentication from working Includes evidence for Anthropic API troubleshooting demand.
copy(credentials=...) carries over inherited api_key even when replacing with credentials provider; AccessTokenAuth short-circuits Fix 429 quota error when using OpenAI Python SDK with ChatGPT Plus OAuth token through third-party apps Includes evidence for OpenAI API troubleshooting demand.
429 quota exceeded for gpt-4 via ChatGPT Plus OAuth third-party Codex Fix ultrareview command consuming free quota even when the command fails due to upstream server-side rate limiting (user did not actually consume the tool) Includes evidence for Claude Code troubleshooting demand.
API Error: Server is temporarily limiting requests (not your usage limit) · Rate limited — Ultrareview free quota still consumed Fix MCP OAuth authentication flow where Claude Code fails to launch browser for Atlassian, DataDog, Slack, SourceGraph integrations on Windows with Git Bash + SSH Includes evidence for Claude Code troubleshooting demand.
mcp auth not automatically launching browser on Windows — no errors, just no activity after clicking Authenticate Understand and fix unexpected weekly usage consumption after Anthropic doubled the 5-hour short-term allowance in a recent Claude Code update Includes evidence for Claude Code troubleshooting demand.
Weekly Limits Are Draining Much Faster After the 5-Hour Limit Update Fix unclear error messages when deploying SolidStart/Vinxi apps to Cloudflare Workers — errors don't mention module scope restriction which is the actual cause of build/deploy failure Includes evidence for Cloudflare troubleshooting demand.
Cloudflare Workers deployment fails with unhelpful errors for setInterval at module scope, Database Pool creation at module scope, Auth instance at module scope Fix LiteLLM rate-limit error categorization where router-side RPM/TPM throttling triggers (deployment-level and model-level) are falsely reported as vendor/third-party rate limits Includes evidence for LiteLLM troubleshooting demand.
litellm_rate_limit errors emitted as vendor_rate_limit — dashboards/callbacks attribute router-side throttles to upstream vendor instead of LiteLLM proxy Fix self-hosted Linux runner permissions error when running GitHub Actions workflows inside containers — checkout step throws permission denied saving state files Includes evidence for GitHub Actions troubleshooting demand.
Error: EACCES: permission denied, open '/__w/_temp/_runner_file_commands/save_state_*' Fix docker-in-docker devcontainer feature regression on Ubuntu 26.04 LTS with Docker Engine 29.5.0 — features that worked before now break Includes evidence for Docker troubleshooting demand.
[docker-in-docker] v2 and v3 feature fails now for previously perfectly working devcontainer configurations Fix Bedrock streaming crash when API returns error SSE frames (rate_limit_error, overloaded_error) — should raise typed APIStatusError instead of AttributeError Includes evidence for Anthropic API troubleshooting demand.
AttributeError: 'NoneType' object has no attribute 'model' — Bedrock error payload routed as completion event in streaming Fix broken MCP server pipe causing tools to become unavailable after 10–20 min idle or active use in Claude Code sessions Includes evidence for Claude Code troubleshooting demand.
Error: No such tool available: mcp__plugin_telegram_telegram__reply | FD 0u unix -> (none) [stdin pipe BROKEN] Fix unknown provider error when routing through OpenRouter/LiteLLM proxy — provider auth chain misconfiguration blocks API calls Includes evidence for LiteLLM troubleshooting demand.
Unknown provider 'openai' Fix connection errors caused by incorrect Azure OpenAI endpoint format in official Python SDK examples Includes evidence for OpenAI API troubleshooting demand.
DNS or TLS error when connecting to Azure OpenAI — example code uses https://example-resource.azure.openai.com/ instead of correct https://example-endpoint.openai.azure.com/ Fix false-positive cybersecurity filter blocking legitimate sysadmin/security-audit work on owned infrastructure; fix session context poisoning that persists across follow-up messages Includes evidence for Claude Code troubleshooting demand.
API Error: Claude Code is unable to respond to this request, which appears to violate our Usage Policy — cyber-related safeguards block routine sysadmin commands and poison entire session Fix VS Code extension ignoring the dangerouslySkipPermissions configuration option — users want automated workflow without repeated permission prompts Includes evidence for Claude Code troubleshooting demand.
dangerouslySkipPermissions setting does not work in VS Code — still prompts for permission even when setting is enabled Fix billing/quota exhaustion handling — users need to distinguish between temporary rate limits (retry) vs. account-level quota exhaustion (switch key/alert) Includes evidence for OpenAI API troubleshooting demand.
InsufficientQuotaError: 429 insufficient_quota response treated as generic RateLimitError Reduce Python function cold start latency on Vercel Serverless Functions — remove unused files from deployment bundle to improve invocation speed Includes evidence for Vercel troubleshooting demand.
FUNCTION_INVOCATION_TIMEOUT or slow cold start on Vercel Python functions due to unnecessary files in bundle (uv installer metadata, unused vendor files)