Error listings
Fix Anthropic Managed Agents API rejecting all Claude models with 'model is not supported' error despite working on standard Messages API with same key Includes evidence for Anthropic API troubleshooting demand.
model is not supported (on agents.create() or sessions.create() via Managed Agents beta API, while same models work on Messages API) Fix Anthropic MCP connector beta returning wrong content block type (tool_use vs mcp_tool_use) causing tool invocation failures in LLM agent tool declarations Includes evidence for Anthropic API troubleshooting demand.
MCP connector: model intermittently emits tool_use instead of mcp_tool_use for tools declared in mcp_toolset with configs.<tool>.enabled: true Identify what Cloudflare services are currently experiencing outages or degradation so affected sites can diagnose root cause and implement mitigation Includes evidence for Cloudflare troubleshooting demand.
Minor Service Outage — Cloudflare page indicator shows 'minor' severity (updated 2026-05-22T18:52:06.999Z) Fix Docker Desktop 4.59.1 crashing on startup on macOS Apple Silicon with JFS service database corruption preventing engine from starting Includes evidence for Docker troubleshooting demand.
service jfs failed: running: invalid database; Docker Desktop crashes at startup after disk resize permission error on Docker.raw Fix OpenAI API Python SDK hanging calls behind NAT/EKS/Gateway environments where TCP keepalive is missing and idle connections get dropped silently Includes evidence for OpenAI API troubleshooting demand.
Non-streaming API calls silently hang forever behind NAT — default httpx transport has no TCP keepalive; connection blocks indefinitely while server generates response Fix Claude Code custom MCP connector failing to persist OAuth tokens across app restarts, showing 'Failed to connect' instead of reconnecting automatically Includes evidence for Claude Code troubleshooting demand.
Authentication successful, but server reconnection failed. You may need to manually restart Claude Code for the changes to take effect. ✗ Failed to connect Fix Cursor IDE Claude Code extension disconnecting immediately after launch, /ide command returns no available IDEs Includes evidence for Cursor troubleshooting demand.
[ERROR] MCP server "ide" Failed to fetch tools: MCP error -32000: Connection closed ... (repeats indefinitely) Fix GitHub Actions workflows failing on pull requests due to repository secrets not being accessible in fork/PR context, while succeeding on direct branch pushes Includes evidence for GitHub Actions troubleshooting demand.
Workflow succeeds on main but fails on PR with 'secret not found' error Fix LiteLLM proxy incorrectly reporting rate limit errors as deployment unavailability, confusing debugging for teams using multi-provider routing Includes evidence for LiteLLM troubleshooting demand.
Rate limit error reported as "No deployments available for litellm proxy" Fix LiteLLM silently ignoring configured timeout values when proxying requests to AWS Bedrock and Google Vertex AI, causing unpredictable request latency Includes evidence for LiteLLM troubleshooting demand.
litellm.Timeout: Connection timed out. Timeout passed=Timeout(timeout=0.1), time taken=0.101 seconds Fix receiving 429 insufficient quota errors on newly activated paid OpenAI accounts that should have unlimited usage Includes evidence for OpenAI API troubleshooting demand.
Error 429: Insufficient Quota - I just signed-up and paid, but got error-429. My limit is still untouched Handle Anthropic API server-overloaded / rate-limit errors in production workflows to avoid blocking LLM-powered applications Includes evidence for Anthropic API troubleshooting demand.
APIResponseError: Request too fast, try again later. Anthropic overloaded_error Fix runaway Claude Code SDK/spawn loops that silently consume Max subscription quota without any per-account rate-limit warnings or visibility controls Includes evidence for Claude Code troubleshooting demand.
Background claude --print spawns silently burn Max5 quota; no per-account rate-limit visibility. Fork-loop dispatcher produces thousands of spawns consuming 84% of Max5 5h-window with no warning. Fix LiteLLM proxy to mask sensitive data (SHA-256 hashes, API keys) in rate limit error responses exposed to end-users Includes evidence for LiteLLM troubleshooting demand.
Rate limit error message body leaks full SHA-256 hash of API key via litellm/proxy/hooks/parallel_request_limiter_v3.py Fix LiteLLM retry mechanism to properly pause/respect configured delay after hitting rate limits, avoiding retry storms that waste API quota Includes evidence for LiteLLM troubleshooting demand.
Retry mechanism not pausing for set delay after rate limit is hit — immediate retries instead of respecting configured backoff window Fix LiteLLM stream_timeout enforcement so timeout errors trigger at correct time rather than being ignored on first chunk, preventing silent hangs in production API calls Includes evidence for LiteLLM troubleshooting demand.
stream_timeout not enforced on first chunk; timeout should raise error at N seconds but actual behavior delays error raising significantly beyond configured timeout Fix Anthropic SDK for Bedrock so error SSE events (rate_limit_error, model_not_found) are properly decoded as error events rather than being misclassified as completion events Includes evidence for Anthropic API troubleshooting demand.
AWSEventStreamDecoder emits every decoded frame as ServerSentEvent(event=completion) even for Bedrock error payloads like rate_limit_error and model-not-found, causing client code to process error events as completions Fix Cloudflare Go SDK DNS deserialization bug where MX records lose their priority field because union type discriminator incorrectly picks ARecord type Includes evidence for Cloudflare troubleshooting demand.
cloudflare-go union type discriminator resolves MX records as RecordResponseARecord instead of RecordResponseMXRecord; r.Priority always returns 0 for MX records despite API returning correct value Fix Next.js 16.2.0 regression causing sitemap.xml builds to fail when using force-static dynamic routing on Vercel; local next build succeeds but cloud deployment fails Includes evidence for Vercel troubleshooting demand.
Error: Invariant: failed to find source route /[store]/[lang]/c/sitemap.xml for prerender /[store]/[lang]/c/sitemap.xml — only reproducible on Vercel remote builder with Turbopack parallel workers Fix OpenAI Python SDK so background-mode Responses API failures raise proper typed exceptions like RateLimitError / InternalServerError instead of opaque free-form error codes Includes evidence for OpenAI API troubleshooting demand.
Responses API background runs return HTTP 200 OK with status="failed" but no typed exception; error.code (e.g. "server_error", "rate_limit_exceeded") does not map 1:1 to any SDK exception class