Unverified50% confidenceCautionExact time
Retries only apply to idempotent requests or operations that can be safely repeated, and calls with side effects should use idempotency keys.
1
Sources
50%
Confidence
Long-term
Relevance
10/25/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedCommon implementations of idempotency include: generating a unique request ID for each tool call and deduplicating in the external system; replacing blind writes with "check-then-write" upsert semantics; or explicitly recording the completion status of each side effect in the checkpoint and skipping completed calls upon recovery.72% similarUnverifiedIf a tool call has already written data to an external system but the checkpoint has not yet been persisted, recovery may result in duplicate writes; durable execution needs to be paired with idempotent design, either making the tool itself idempotent or recording a completion marker for the side effect in the checkpoint.69% similarUnverifiedWhen accepting output-type tasks, one must not only check whether the final state meets the conditions but also confirm that this request indeed triggered a new write or creation, because idempotent systems make repeated execution produce no new changes.69% similarUnverifiedDeclare retry policies at the node level, including maximum retry counts, backoff strategies and retryable exception types; combined with type-checking code, retry configurations can be verified at compile time, avoiding the discovery of missing policies only at runtime.69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1096163API
curl https://kongchang.com/api/v1/knowledge/claims/1096163MCP
get_claim(id=1096163)