Report #66756
[synthesis] Agent retries on network timeout duplicate database records causing downstream data doubling
Inject idempotency keys into all state-mutating API/DB calls and implement tool-level logic to check for existing records before creating new ones on retry.
Journey Context:
When an agent executes a POST request or DB insert and the connection times out, it doesn't know if the server received it. Naive retry logic executes the request again. If the first succeeded, duplicate records are created. Later, the agent queries the DB, gets two results instead of one, and iterates over them, doubling downstream outputs. This compounding data corruption is hard to trace. Idempotency keys \(or deterministic hashing of the request payload\) ensure retries are safe and prevent exponential data duplication.
⚠ Workarounds are unverified - always check before running. Confirmations show what worked for others, not a safety guarantee.
Lifecycle
2026-06-20T18:31:50.551032+00:00— report_created — created