completion-marker-enforcement

Orchestrator detection fails when completion markers are output incrementally during character-by-character streaming. Enforcing strict protocol compliance ensures the marker is sent as an atomic, single string at the response end, preventing parsing issues.

To format completion markers correctly, output the marker as a single atomic string at the very end of the assistant response. Protocol enforcement validates the response structure to prevent shell commands or tool calls after the marker.

Protocol enforcement for task completion markers guarantees strict compliance for outputting designated tags like `<ralph>COMPLETE</ralph>`. It provides pre-output validation to prevent orchestrator detection failures and downstream processing errors.

Executing tool calls after outputting a completion marker violates protocol enforcement. Validation patterns ensure no shell commands or tool calls occur after the marker, guaranteeing the marker remains the final output to prevent parsing issues.

To stop incremental output from breaking detection, apply system workarounds that enforce atomic, single-string output of the completion marker. This mitigates character-by-character streaming issues and ensures proper orchestrator detection.

Yes, completion marker validation works without external dependencies. The enforcement mechanism uses internal references to validate response structures and ensure protocol compliance natively, requiring no additional environment setup.