Read MCP results without mistaking transport success for tool success
Separate HTTP, JSON-RPC and operational failures, then report only the result the server established.
Lupin field guides
Articles 145–162 of 200
Separate HTTP, JSON-RPC and operational failures, then report only the result the server established.
Collect allowed public records with explicit page identity, deduplication and a finite budget.
Locate the boundary between recognized words and a completed caller turn before tuning voice latency.
Inspect durable meter exports, stable identifiers and invoice guards when accepted usage has not reached the billing provider.
Use durable checkpoints and provider evidence to recover a paid action whose response was lost, while preserving the original operation identity.
Preserve command identities, inspect history and artifacts, and distinguish API deduplication from application idempotency.
Distinguish inactive grants, insufficient scope and provider outages while preserving the replacement authorization lifecycle.
Handle lost responses, conflicting requests and revoked credentials without creating replacement workspaces.
Distinguish link lifetime, account activation and uncertain creation before deciding whether to reconnect.
Triage a provider incident by capability, admission, known outcomes and cleanup before releasing a queued backlog.
Read stored, reserved and deleting bytes separately so content cleanup cannot advertise space that is still committed or unresolved.
Design bounded structured diagnostics that preserve operation evidence while excluding credentials, private content and raw provider errors.
Create a bounded Free workspace and inbox without a human account, while keeping registration credentials recoverable.
Build status tracking around actual event evidence rather than treating a successful send response as a read receipt.
Investigate changed request data, serialization and operation identity when a durable action returns an idempotency conflict.
Distinguish authentication blueprints, provider credentials and customer consent when an integration is unavailable.
Classify validation, permission, provider and uncertain-action failures before deciding what can be retried.
Retain operation identity, concrete arguments and provider evidence across timeouts and durable recovery.