Resume and govern an autonomous RLXP experiment loop from existing study state.
npx skills add https://github.com/junhyekh/rlxp --skill rlxp-autoloop
Please help me install this Agent Skill. Skill: rlxp-autoloop Source: https://github.com/junhyekh/rlxp/tree/main/plugins/rl-experiment-assistant/skills/rlxp-autoloop Command: npx skills add https://github.com/junhyekh/rlxp --skill rlxp-autoloop
This Skill helps you resume and automate an approved reinforcement-learning experiment loop without mixing tasks, violating scope, or launching unsafe runs.
.rlxp/ state, contract, ledger, reports, candidates, and run artifacts before deciding the next action.Ask the assistant to resume the RLXP autoloop for the active task and continue iterating until the approved budget or stop rule is reached.
High-intent search queries and answers about installing and using this skill.
Automating a reinforcement learning experiment loop safely requires validating launch gates, enforcing budget and scope checks, and executing training only through a bundled autoloop runner to preserve deterministic state updates.
State management for resuming an RL experiment works by reading the existing `.rlxp/` state, contract, ledger, reports, candidates, and run artifacts to decide the next action without mixing scopes.
Validating RL candidates and enforcing a training budget automatically is possible through budget gating, candidate lifecycle tracking, and contract checks applied before any training run launches.
Automated RL experiment control requires an approved reinforcement-learning task and study repository with existing state, ensuring the loop resumes from established artifacts rather than starting fresh.
RL training runs fail launch validation when contract checks, budget gates, or scope boundaries are violated, preventing unsafe runs from executing outside the approved experiment parameters.
Comparing reinforcement learning run results against an incumbent involves evidence-driven decisions where the loop classifies outcomes, proposes the next candidate, and updates study artifacts based on the comparison.
junhyekhConvert RL experiment ideas into scoped contracts with metrics and approval gates.
product-on-purposeDocument A/B test results with metrics, learnings, and recommendations.
junhyekhCoordinate reinforcement-learning experiment loops with validation, monitoring, and launch gates.
junhyekhCreate a study-scoped reinforcement-learning baseline plan with training commands and metric paths.