Re-execute the Algorithm in loop mode to iteratively refine a target.
npx skills add https://github.com/julianobarbosa/.claude --skill loop-julianobarbosa
Please help me install this Agent Skill. Skill: Loop Source: https://github.com/julianobarbosa/.claude/tree/main/skills/Loop Command: npx skills add https://github.com/julianobarbosa/.claude --skill loop-julianobarbosa
The Loop skill helps you iteratively refine a target by running multiple full algorithm passes, so each revision builds on the previous one toward an ideal outcome instead of a one-shot attempt.
Run the loop by telling it which file or directory to improve and how you want it to get better, for example: loop the Algorithm on the file and improve it for consistency over 5 iterations.
High-intent search queries and answers about installing and using this skill.
Iterative document refinement is achieved by re-executing a full algorithm cycle on a fixed target, allowing each revision to build on the previous one with accumulated context. This process runs multiple passes toward an ideal outcome instead of a single attempt.
A human-in-the-loop workflow inserts review and approval phases between iterative cycles. This ensures that automated improvement changes remain aligned with your judgment and goals throughout the repeated algorithm passes.
Yes, you can enable an opt-in autoresearch mode for autonomous iteration. This mode uses a dead-ends ledger and confidence gating based on measurable iteration scores to manage the workflow without manual review between cycles.
To start an iterative improvement loop, specify the target file or directory and define the improvement goal. For example, instruct the system to loop the algorithm on the file to improve consistency over a set number of iterations.
Workflow optimization through repeated cycles requires tracking the iteration count and evolving the evaluation criteria per pass. This OBSERVE→LEARN feedback approach ensures each cycle improves upon the last based on measurable outcomes.
starlink-awakenAutomates iterative Algorithm cycles with human review between each pass.
davduncExecute multiple Algorithm cycles with iteration-to-iteration learning and review.
bigeasyfreemanOrchestrate multi-step workflows through an ISC framework with phase-driven execution.
PaulJPhilpImplement Effect-TS scheduling patterns for periodic tasks with cron, debounce, and retry.
