TR

train-with-environments

Helps configure and troubleshoot RL training runs using prime-rl and native v1 tasksets.

Install

mkdir -p .claude/skills/train-with-environments && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/4067" && unzip -o skill.zip -d .claude/skills/train-with-environments && rm skill.zip

Installs to .claude/skills/train-with-environments

Activation

This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.

Configure and diagnose prime-rl training with native verifiers.v1 tasksets, harnesses, runtimes, and trace branches. Use for RL experiment setup, environment-server configuration, group sizing, difficulty filtering, renderer selection, periodic eval, or rollout/training failure diagnosis.
289 chars✓ has a “when” triggerlonger than Claude Code's old 250-char listing cap (fine on current versions)
Advanced

Key capabilities

  • Configure `prime-rl` training with native verifiers.v1 tasksets
  • Diagnose rollout failures separately from training instability
  • Set up difficulty filtering for training environments
  • Configure renderers and handle trace branches
  • Manage multi-environment training with sampling ratios

How it works

This skill configures `prime-rl` to train models using native v1 tasksets, allowing for detailed control over environment shape, group sizing, and difficulty filtering.

Inputs & outputs

You give it
Validated v1 taskset and `prime-rl` configuration settings
You get back
Prime-RL version/source, resolved config, pretraining validation evidence, and failure classifications

When to use train-with-environments

  • Configure reinforcement learning training runs
  • Diagnose rollout and training failures
  • Tune hyperparameters for RL agents

About train-with-environments

Assists developers in setting up training loops, tuning hyperparameters, and diagnosing instability in RL experiments. It provides a standardized way to integrate native tasksets into training workflows.

Train models with verifiers environments using hosted RL or prime-rl. Use when asked to configure RL runs, tune key hyperparameters, diagnose instability, set up difficulty filtering and oversampling, or create practical train and eval loops for new environments.

When not to use it

  • When the taskset has not been validated for correct prompt/tool/user traces and reward
  • When the hosted training's public CLI schema does not support native v1 taskset/harness fields
  • When the project does not use `prime-rl` for reinforcement learning

Limitations

  • The skill requires `prime-rl` to be installed and configured.
  • The skill assumes the use of `vf.Trace` over the v1 environment-server protocol.
  • The skill does not support native v1 hosted commands until the public CLI schema changes.

How it compares

This skill provides a structured approach to integrating and diagnosing v1 tasksets within `prime-rl`, unlike a generic RL setup that might not differentiate between rollout and training instability.

Compared to similar skills

train-with-environments side by side with the closest alternatives in the catalog.

SkillInstallsUpdatedSafetyDifficulty
train-with-environments (this skill)126dReviewAdvanced
senior-data-scientist97moReviewAdvanced
llama-cpp218moReviewIntermediate
langchain268moReviewIntermediate

Try saying

Example prompts that trigger this skill in your AI assistant.

You might also like

senior-data-scientist

davila7

World-class data science skill for statistical modeling, experimentation, causal inference, and advanced analytics. Expertise in Python (NumPy, Pandas, Scikit-learn), R, SQL, statistical methods, A/B testing, time series, and business intelligence. Includes experiment design, feature engineering, model evaluation, and stakeholder communication. Use when designing experiments, building predictive models, performing causal analysis, or driving data-driven decisions.

952

llama-cpp

zechenzhangAGI

Runs LLM inference on CPU, Apple Silicon, and consumer GPUs without NVIDIA hardware. Use for edge deployment, M1/M2/M3 Macs, AMD/Intel GPUs, or when CUDA is unavailable. Supports GGUF quantization (1.5-8 bit) for reduced memory and 4-10× speedup vs PyTorch on CPU.

21471

langchain

zechenzhangAGI

Framework for building LLM-powered applications with agents, chains, and RAG. Supports multiple providers (OpenAI, Anthropic, Google), 500+ integrations, ReAct agents, tool calling, memory management, and vector store retrieval. Use for building chatbots, question-answering systems, autonomous agents, or RAG applications. Best for rapid prototyping and production deployments.

26138

llama-factory

zechenzhangAGI

Expert guidance for fine-tuning LLMs with LLaMA-Factory - WebUI no-code, 100+ models, 2/3/4/5/6/8-bit QLoRA, multimodal support

15112

llava

zechenzhangAGI

Large Language and Vision Assistant. Enables visual instruction tuning and image-based conversations. Combines CLIP vision encoder with Vicuna/LLaMA language models. Supports multi-turn image chat, visual question answering, and instruction following. Use for vision-language chatbots or image understanding tasks. Best for conversational image analysis.

7117

cocoindex

cocoindex-io

Comprehensive toolkit for developing with the CocoIndex library. Use when users need to create data transformation pipelines (flows), write custom functions, or operate flows via CLI or API. Covers building ETL workflows for AI data processing, including embedding documents into vector databases, building knowledge graphs, creating search indexes, or processing data streams with incremental updates.

6116

Search skills

Search the agent skills registry