Skill 与 AgentSKILLS & AGENTS
redai-studio/Relax
An Asynchronous Reinforcement Learning Engine for Omni-Modal Post-Training at Scale
根据公开文档整理，未运行实测；费用以项目官网为准。Based on public documentation, without runtime testing. Check the project site for pricing.
能力CAPABILITIES
上手GET STARTED
docker pull ghcr.io/redai-studio/relaxrl:latest选择与你的环境对应的方式。命令来自项目文档，未运行实测。Choose the option for your environment. Commands come from project documentation and have not been run-tested.
使用限制LIMITATIONS
以下要求与风险来自文档静态检查，未执行项目代码；扫描没有命中，不代表使用安全。Risk here is evidence, not an alarm. These findings come from static scanning of the docs; we never execute a project’s code.
限制LIMITS
需检查Needs review
静态扫描STATIC SCAN
01执行 Shell 命令Runs shell commandsskills/debug-hang/SKILL.md
allowed-tools: bash
02读写本地文件Reads and writes local filesskills/nemo-gym-recipe-integration/SKILL.md
2. Write the raw JSONL directly to a filesystem visible at the same absolute path from every remote Ray node.
03读取 API KeyReads an API keyskills/ssh-ray-cluster/SKILL.md
Read SSH credentials and `RELAX_PROJECT_ROOT` from auto-memory (`reference_ray_cluster_ssh.md`). Ask the user if missing — never hard-code i
项目动态PROJECT ACTIVITY
首次发现FIRST SEEN
26 OCT 202626 OCT 2026
状态STATUS
持续活跃Active
依赖：Dependencies:markdown-it-mathjax3
{
"agents": [
{
"agent": "claude-code",
"evidence": "tree: .claude/"
},
{
"agent": "codex",
"evidence": "tree: .codex/"
},
{
"agent": "claude-ai",
"evidence": "readme: …set of [Claude Code](https://claude.ai/code) slash-command skills un…"
},
{
"agent": "opencode",
"evidence": "skills/creating-skills/SKILL.md: …dev/ │ └── git-commit/ ├── .opencode/ │ └── skills -> ../skills…"
}
],
"apiKey": "api_key = os.environ.get(\"DEEPEYES_JUDGE_API_KEY\") or os.environ.get(\"OPENAI_API_KEY\")",
"docker": null,
"taxonomy": {
"scores": {
"design": 0,
"ai-tools": 1,
"dev-tools": 0,
"indie-web": 0,
"skill-agent": 0,
"productivity": 0
},
"primary": "skill-agent",
"secondary": null
},
"localRuntime": "Advantages run on independent GPU clusters with streaming data exchange a",
"skillMdTotal": 19,
"skillMdErrors": [],
"treeTruncated": false,
"categoryScores": {
"data": 2,
"agent": 7,
"media": 1,
"design": 0,
"browser": 0,
"devtool": 0,
"security": 0,
"marketing": 0,
"productivity": 0
},
"classification": [
{
"signal": "19 SKILL.md with name + description",
"weight": 0.95
},
{
"signal": "skills/ directory with 55 markdown files",
"weight": 0.75
},
{
"signal": "MCP SDK dependency",
"weight": 0.6
}
],
"scannedSources": [
"readme",
"skills/agentic-rollout/SKILL.md",
"skills/code-review/SKILL.md",
"skills/create-training-script/SKILL.md",
"skills/creating-skills/SKILL.md",
"skills/debug-hang/SKILL.md",
"skills/dev/SKILL.md",
"skills/doc-writer/SKILL.md",
"skills/git-commit/SKILL.md",
"skills/model-integration/SKILL.md",
"skills/nemo-gym-recipe-integration/SKILL.md",
"skills/opd-tuning/SKILL.md",
"skills/perf-doctor/SKILL.md",
"skills/redaccel-to-relax/SKILL.md",
"skills/relax-github-ci/SKILL.md",
"skills/relax-github-workflow/SKILL.md",
"skills/sglang-upgrade/SKILL.md",
"skills/ssh-ray-cluster/SKILL.md",
"skills/sync-github/SKILL.md",
"skills/verl-to-relax/SKILL.md"
],
"skillMdFetched": 19
}