Implement code changes and iterate with a persistent reviewer until zero findings remain.
npx skills add https://github.com/Idate96/codex_skills --skill ralpha-loop
Please help me install this Agent Skill. Skill: ralpha-loop Source: https://github.com/Idate96/codex_skills/tree/main/skills/ralpha-loop Command: npx skills add https://github.com/Idate96/codex_skills --skill ralpha-loop
This Skill automates a rigorous implement-test-review cycle for coding tasks, ensuring code quality and adherence to requirements through strict, iterative feedback.
Use the ralpha-loop skill to implement the requested code changes and iterate with the reviewer until no findings remain.
High-intent search queries and answers about installing and using this skill.
To automate the code review and testing loop, this Skill implements a deterministic process that writes code, runs tests, and requests review, repeating until the reviewer has zero findings.
A persistent reviewer agent is reused across all iteration cycles to maintain context and consistency. It improves iterative coding by systematically tracking findings in a structured ledger until all issues are resolved.
You ensure code quality by using an automated implement-test-review cycle that requires a persistent reviewer agent and a structured findings ledger to manage strict iteration until the reviewer is satisfied.
Yes, a structured findings ledger is required for effective cycle management. It systematically tracks and addresses reviewer feedback during the implement-test-review loop to ensure all findings are resolved.
Your automated coding loop may fail the review phase if the reviewer agent identifies unresolved findings. The loop is designed to repeat implementation, testing, and review until the reviewer reports zero findings.
The best way to manage reviewer feedback is through a structured findings ledger that tracks issues systematically, allowing the implement-test-review loop to address all findings before completing the coding task.
robertpelloniCoordinate 3+ independent failure investigations by dispatching parallel agents.
abdias98Implement minimal production code to pass failing tests with checkpoint commits.
macaughDetects injection, memory safety and authorization flaws in multiple languages.
levnikolaevichOrchestrate two-pass Story quality reviews across code, regression, and manual testing.
