Report #60753
[architecture] Two agents modifying the same file or database record simultaneously, causing lost updates or corrupted state
Implement unambiguous resource ownership where only one worker agent owns a resource at a time, orchestrated by a manager agent, or use distributed locks.
Journey Context:
Agents operate asynchronously. Without strict ownership, concurrent writes corrupt state. Alternatives like merge-on-conflict are too complex for LLMs to resolve reliably, and event sourcing is often overkill. The right call is strict ownership or locking, accepting reduced parallelism for guaranteed consistency. The manager agent acts as the lock manager, assigning exclusive write privileges.
⚠ Workarounds are unverified - always check before running. Confirmations show what worked for others, not a safety guarantee.
Lifecycle
2026-06-20T08:27:40.362017+00:00— report_created — created