omc-ralph

Continuous verification loops for software development tasks run a persistent execution cycle, requiring verification until successful completion or a failure limit is reached. The loop identifies verification failures and prompts for iterative fixes.

A persistent task execution loop is a verification cycle that keeps running until a task passes all checks or hits a hard blocker. You need it for quality assurance scenarios like repeatedly testing and fixing a software component until it passes.

Iterative fixing for a failing build task is handled by triggering the continuous verification loop on the target task. If verification fails, the loop identifies the specific issue and prompts for a fix before attempting the execution again.

Continuous iteration works for end-to-end feature testing by applying the verification loop directly to the build feature task. It executes the task persistently, running checks and prompting for fixes until all tests pass or the failure limit is reached.

When a verification loop reaches its failure limit, the persistent execution stops. This mechanism prevents infinite loops by halting the continuous iteration when a hard blocker is encountered or the maximum allowed verification failures are exceeded.

A persistent verification loop automates the repetitive cycle of task execution and checking, unlike manual task execution. It continuously iterates and prompts for fixes until successful completion, ensuring quality assurance without constant manual intervention.