Cause (Documented platform behavior): _use_aiohttp() returns True when aiohttp is importable and the caller did not pass an httpx async client or transport. Async requests then go through aiohttp with timeout=aiohttp.ClientTimeout(total=http_request.timeout) (None when HttpOption
Cause (Documented platform behavior): 'timeout server' is an inactivity timeout that in HTTP mode effectively bounds the server's processing time before headers; 'timeout tunnel' overrides client/server timeouts for tunnels (e.g. WebSocket). Termination state sH = timeout server
Cause (Documented platform behavior): Local mode takes an exclusive file lock on the storage directory; it does not support concurrent access. Fix status: documented_behavior Misleading approaches: - Deleting the lock file while another process still uses the folder risks corrupt
Cause (Maintainer-confirmed cause): Messages were stringified with full object representations (metadata included) into the summary prompt, inflating tokens. Fix status: fixed_upstream Workaround (not a fix): Custom summarization that formats only role + content. Limitations: - O
Cause (Documented platform behavior): System.ClientModel's ClientPipeline.DefaultNetworkTimeout is 100 s (per try) unless ClientPipelineOptions.NetworkTimeout is set; Azure.Core RetryOptions.NetworkTimeout defaults to 100 s; HttpClient.Timeout defaults to 100 s and reports 'The r
Cause (Documented platform behavior): httpx checks for the h2 package when http2=True and raises ImportError with the install hint. Fix status: documented_behavior Limitations: - httpx2 equivalent message not separately verified. Evidence (public sources, summarized; not reproduc
Cause (Documented platform behavior): The Tools Agent binds tools through the chat model interface. Fix status: documented_behavior Evidence (public sources, summarized; not reproduced by this contributor): - https://raw.githubusercontent.com/n8n-io/n8n/2f2bac4645472e5efc72c4efed
Cause (Documented platform behavior): n8n wraps LangChain parser errors into this NodeOperationError with the On Error hint. Fix status: documented_behavior Limitations: - Auto-fixing parser suggestion based on node availability, not a doc statement in this source. Other error fr
Cause (Documented platform behavior): SDK validates against get_gpu result. Fix status: documented_behavior Limitations: - Example ID format is illustrative. Evidence (public sources, summarized; not reproduced by this contributor): - https://raw.githubusercontent.com/runpod/runp
Cause (Documented platform behavior): Auto prompt mode expects the Chat Trigger output field chatInput (guardrails mode expects guardrailsInput); define mode errors when the expression resolves to undefined. Fix status: documented_behavior Other error fragments: - Expected to fin
Cause (Documented platform behavior): vLLM derives a maximum length from the HF config and refuses larger values unless explicitly overridden, since positions beyond it produce NaNs (RoPE) or out-of-bounds errors (absolute positions). Fix status: documented_behavior Misleading ap
Cause (Documented platform behavior): Graph validation requires a single entry and (for sequential agents) a terminal End/Loop node. Fix status: documented_behavior Limitations: - "orphan nodes count as starting nodes" inferred from starting-node computation. Other error fragment
Cause (Documented platform behavior): undici implements the Fetch spec Request constructor: step 34 throws when GET/HEAD has a body; step 38 throws when the body has no fixed source (a stream) and init.duplex is missing (only 'half' is supported). Fix
Cause (Documented platform behavior): Default maxRetries=2, initialDelayInMs=2000, backoffFactor=2. isRetryableStatusCode treats 408, 409, 429 and >=500 as retryable. getRetryDelayInMs uses retry-after-ms or retry-after only when 0 <= ms and (ms < 60 s or ms < the backoff delay).
Cause (Documented platform behavior): 401 mapped to AuthenticationError. Fix status: documented_behavior Evidence (public sources, summarized; not reproduced by this contributor): - https://raw.githubusercontent.com/runpod/runpod-python/760aea2cc6c0a8f5739b2a8667f61ae11d49d37b/ru
Cause (Documented platform behavior): vLLM guided decoding can apply only one constraint; a named tool_choice itself is implemented as a JSON constraint, so both cannot be combined. Fix status: documented_behavior Limitations: - Mechanism (named tool_choice as guided JSON) is inf
Cause (Documented platform behavior): Local mode runs one job from test_input.json / --test_input; non-zero exit on missing input or job error (documented for CI). Fix status: documented_behavior Limitations: - Local-mode trigger condition (env var) not verified in the fetched li
Cause (Documented platform behavior): BaseRequestsTool.__init__ raises unless allow_dangerous_requests is truthy, because model-controlled URLs can reach internal servers (SSRF); the message recommends a proxy and sandboxing. Fix status: documented_behavior Misleading approaches:
Cause (Documented platform behavior): Key resolution order: FLOWISE_SECRETKEY_OVERWRITE, AWS Secrets Manager, then key file; if the file cannot be read a new random key is written. Fix status: documented_behavior Misleading approaches: - Re-entering API keys in the flow node with