Error listings
Debug why Anthropic SDK with MCP beta returns plain tool_use instead of expected mcp_tool_use, causing agent loops to hang 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 — conversation stalls on client side Fix Claude Code 400 invalid_request_error when Claude attempts to download an image (from Notion), upload it elsewhere (to LinkedIn), and then chokes — blocking all subsequent messages Includes evidence for Claude Code troubleshooting demand.
API Error: 400 {"type":"error","error":{"type":"invalid_request_error","message":"Could not process image"}} Fix OAuth 2.1/PKCE auth flow for remote MCP servers where Claude Code reports success but tools never become available Includes evidence for Claude Code troubleshooting demand.
MCP OAuth completes but token is not honored — server stays in 'Needs authentication' Understand why Cloudflare Response Header Transform Rules don't apply to WAF challenge block responses, breaking cross-origin JavaScript fetch() calls Includes evidence for Cloudflare troubleshooting demand.
WAF challenge 403 response lacks Access-Control-Allow-Origin CORS headers despite Response Header Transform Rule configured — cross-origin fetch() calls fail Resolve Vercel deployment build failures caused by stricter TypeScript type checking compared to local tsc/tsup builds Includes evidence for Vercel troubleshooting demand.
vercel build TypeScript errors: Property 'body' does not exist on type 'Response' — vercel build produces TS errors even when tsc/tsup succeed locally Debug why Cursor cloud agent falls back to default base image after environment configuration fails; fix Dockerfile-based env setup with build secrets Includes evidence for Cursor troubleshooting demand.
Environment configuration failed — agent defaults to base image with warning signs instead of running in configured environment Resolve secondary rate limit error in GitHub Actions workflow caused by excessive concurrent API calls; understand point calculation for secondary limits Includes evidence for GitHub Actions troubleshooting demand.
Secondary rate limit triggered: too many concurrent requests (exceeds 100 concurrent limit) or 900 points/minute REST endpoint threshold Fix blocked deployment when Vercel marks project/team as blocked due to fair use violation; restore ability to deploy Includes evidence for Vercel troubleshooting demand.
Blocked scope returns HTTP 451 error; user receives "Team access required to deploy." message with manual approval gate Fix invalid route source pattern error in vercel.json where negative lookahead regex `(?!pattern)` doesn't work — must use `((?!pattern).*)` syntax with path-to-regexp Includes evidence for Vercel troubleshooting demand.
Invalid route source pattern — path-to-regexp syntax vs RegExp confusion for negative lookaheads in vercel.json Fix confusing 400 errors when uploading skills to Anthropic API - first error is wrong, second error is still misleading Includes evidence for Anthropic API troubleshooting demand.
SKILL.md file must be exactly in the top-level folder. Fix docker compose pull silently skipping volume images so containers run stale image versions Includes evidence for Docker troubleshooting demand.
docker compose pull does not update images used in volumes with type=image Fix large file proxy failures above 100KB returning partial content instead of full response in Kong Kubernetes deployments Includes evidence for Kubernetes troubleshooting demand.
net::ERR_HTTP2_PROTOCOL_ERROR 206 Partial Content; recv buffer stuck with ZeroWindow flag Fix Next.js notFound() prerender causing broken page hydration and flight data holes with cacheComponents enabled Includes evidence for Vercel troubleshooting demand.
notFound() prerender leaves holes in Flight data that never resolve resulting in reported hydration errors Fix Next.js route cache serving stale data when navigating between URLs with different repeated search params like color=red&color=green vs color=red&color=blue Includes evidence for Vercel troubleshooting demand.
client-side cache key collision for repeated search param keys results in stale render, Suspense never re-suspends Fix 400 error when deploying gpt-image-1.5 on Azure OpenAI with mismatched deployment and model names Includes evidence for OpenAI API troubleshooting demand.
Model not supported with Responses API. Supported models are: ['gpt-image-1', 'gpt-image-1-mini', 'gpt-image-1.5'] Fix unhandled async/Background response failures in openai-python SDK where errors surface as opaque strings without typed exception classes Includes evidence for OpenAI API troubleshooting demand.
Background responses API returns status='failed' with error.code='server_error' via HTTP 200 — no typed exception raised; caller cannot catch RateLimitError or InternalServerError Fix Claude Code MCP authentication failure when connecting to self-hosted instance where oauth2 authorize rejects offline_access scope then callback gives 'requested resource invalid' Includes evidence for Claude Code troubleshooting demand.
error=invalid_scope error_description='The following scopes are invalid: offline_access'; subsequent auth gives 'requested resource invalid' Fix Docker Sandboxes sbx run failing on Windows 11 because sandboxd in-process moby backend cannot connect to its own socket immediately after starting Includes evidence for Docker troubleshooting demand.
ERROR: failed to create sandbox: create runtime: sandboxd error: status 500: failed to create network: Cannot connect to the Docker daemon at unix://...docker.sock. Is the docker daemon running? Fix GitHub Copilot authentication failures across multiple editors (Neovim/copilot.vim, Zed) on macOS where auth endpoint returns malformed responses despite valid TLS Includes evidence for GitHub Copilot troubleshooting demand.
E-32603 Response content-type is missing (status=200); HTTP 200 response does not appear to originate from GitHub Fix LiteLLM router strategy incorrectly attributing self-imposed rate limits (deployment over defined rpm/tpm limit) to upstream vendors, preventing correct dashboards/alerts for proxy-side throttling Includes evidence for LiteLLM troubleshooting demand.
RateLimitError silently emitted category='vendor_rate_limit' for router-side rpm/tpm enforcement instead of category='litellm_rate_limit'