Dynamic Reassignment Skill

Automated agent reassignment switches failed tasks to alternative agents defined in fallback chains. It handles persistent errors like verification failures or crashes by delegating the task to a different agent, preserving progress through checkpointing before all options are exhausted.

Rate limit handling with exponential backoff manages transient errors by waiting and retrying the agent task before attempting reassignment. This approach resolves temporary API rate limits without switching agents, avoiding unnecessary context loss during delegation.

Context overflow recovery handles token limits by creating a checkpoint of the current progress and starting a fresh agent session. This preserves the task state and allows the alternative agent to continue execution without losing prior work.

When all fallback agent reassignment attempts are exhausted, the system follows an escalation protocol that notifies the user for intervention. This ensures no task is silently dropped and provides a clear path for manual resolution.

Yes, you can define custom fallback chains to switch to specific alternative agents based on failure types. For example, if a tester agent fails verification, the task automatically reassigns to a developer agent before escalating to the user.

Dynamic task delegation is the process of automatically reassigning tasks to alternative agents upon failure. You need it when running multi-agent workflows where transient rate limits, context overflows, or verification failures could otherwise halt task completion.