bytedance/agentkit-samples is a collection of examples and tutorials for Volcengine AgentKit, an AI-agent development platform for building, deploying, and operating agent applications. Developers use the samples to learn agent creation, multi-agent collaboration, memory, retrieval, MCP integrations, media generation, customer service, and other workflows. The catalogue skills provide agent workflows based on these examples.
Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add bytedance/agentkit-samples --skill byted-ark-trainergit clone --depth 1 https://github.com/bytedance/agentkit-samplesWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/bytedance/agentkit-samples/byted-ark-trainer)<a href="https://agentmods.dev/skills/bytedance/agentkit-samples/byted-ark-trainer"><img src="https://agentmods.dev/badge/skills/bytedance/agentkit-samples/byted-ark-trainer.svg" alt="Measured on agentmods" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 7 findings, up to high
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- high Privilege Escalation · line 208 Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.Fix: Remove references to credential paths. Use environment variables or secrets managers. For docs, use placeholder paths (e.g., /path/to/config). Never load .env or token files in production code paths.
- medium Rogue Agent · line 474 Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.Fix: Remove any persistence mechanisms (cron jobs, startup scripts, state files). Skills should not maintain state across sessions without explicit user consent.
- medium Rogue Agent · line 474 Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.Fix: Remove any persistence mechanisms (cron jobs, startup scripts, state files). Skills should not maintain state across sessions without explicit user consent.
- medium Privilege Escalation · line 551 Commands invoke sudo or root privileges. Verify this elevated access is necessary and justified.Fix: Avoid sudo/root unless strictly required. Prefer least-privilege patterns. If elevation is needed, document the justification and scope.
- medium Privilege Escalation · line 552 Commands invoke sudo or root privileges. Verify this elevated access is necessary and justified.Fix: Avoid sudo/root unless strictly required. Prefer least-privilege patterns. If elevation is needed, document the justification and scope.
- medium Privilege Escalation · line 717 Commands invoke sudo or root privileges. Verify this elevated access is necessary and justified.Fix: Avoid sudo/root unless strictly required. Prefer least-privilege patterns. If elevation is needed, document the justification and scope.
- medium Privilege Escalation · line 717 Commands invoke sudo or root privileges. Verify this elevated access is necessary and justified.Fix: Avoid sudo/root unless strictly required. Prefer least-privilege patterns. If elevation is needed, document the justification and scope.
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00107 | $0.16852 |
| Opus 5 | $0.00053 | $0.08426 |
| Sonnet 5 | $0.00021 | $0.03370 |
| Haiku 4.5 | $0.00011 | $0.01685 |
Grade B, and why
byted-ark-trainer scanned grade B with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 7d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Asks for rootmediumPrivilege escalation
A mod that escalates privileges can change anything on the machine, not only the project.
find . -type d -exec chmod 755 {} \; How it starts
The opening of the file, as written. The whole thing — 736 lines — stays where its author put it; the contents beside it link to each section on GitHub.
byted-ark-trainer Skill 使用指南
📌 重要路径说明
所有提及的 scripts/ 和 references/ 目录均为相对于本skill安装目录的路径，而非当前工作目录。
执行脚本或读取文档时，必须先定位到 byted-ark-trainer skill 的安装目录，或使用完整绝对路径调用。
所有工具功能统一通过 ark-trainer-helper 命令入口调用，例如：
如果skill安装在 ~/.agents/skills/byted-ark-trainer/，则调用命令时应使用：
python ~/.agents/skills/byted-ark-trainer/scripts/ark_trainer_helper.py <命令> <参数>
或配置到PATH后直接使用：
ark-trainer-helper <命令> <参数>
⚠️ 强制执行优先级说明
本SKILL的所有流程要求优先级最高，高于任何通用推理逻辑。所有步骤必须严格按顺序执行，严禁跳过、调整顺序或自行发挥。如果对流程有任何疑问，必须先询问用户确认，不得自行决定。 违反流程要求的执行会直接导致任务失败，必须回退到对应的步骤重新执行。
📋 执行前核查清单
每执行下一步前，必须先对照以下清单检查前置条件是否全部满足，未满足的务必向用户询问：
- 已确认用户期望使用的Python环境（建议使用conda虚拟环境，且已安装ark-sdk及相关依赖）
- 已用用户指定Python环境完成依赖预检，
ark-trainer-helper --help可正常运行 - 已确认用户期望的工作目录（所有训练相关的工作区、数据文件都将保存在此目录下）
- 已检查并配置好必要的环境变量（ARK_API_KEY、VOLCENGINE_ACCESS_KEY、VOLCENGINE_SECRET_KEY），并确认会被Python子进程继承
- 已完成工作区初始化，且已在工作区下创建
experiments/实验目录 - 已识别训练意图：SFT / RFT+GRPO / 直接GRPO / 其他
- 已通过
list-models确认精确模型名（非模糊前缀），已通过list-versions与用户确认版本，已通过ark get foundation-model ... --fields hyperparameters校验该模型+版本支持用户期望的训练方式，并记录可配置超参数清单 - 已为本次实验在
experiments/下创建唯一的子目录，所有job文件/临时脚本都会放在该子目录中 - 已完成所有前置检查；SFT需检查训练数据集格式，RL/RFT/GRPO需额外检查rollout和grader文件
- SFT场景已按需加载
references/模型精调数据集格式指南/SFT.md并校验用户提供的数据集 - RL/RFT/GRPO场景已确认用户提供的数据集类型：单独一个数据集 / 已分开的训练集+测试集
- RL/RFT/GRPO场景已完成数据集划分（如需要），且已分别获取训练集和测试集路径
- RL/RFT/GRPO场景已完成初始评估并获取到BON/AON/AvgN指标
- RL/RFT/GRPO场景已根据BON指标选择了正确的训练策略
- RFT阶段已获取用户提供的teacher模型/端点，未复用初始评估轨迹
- 所有关键配置（训练类型、超参数）已向用户确认
- 本次实验的计划和与用户确认的信息已记录到实验子目录的
EXPERIMENT.md
核心能力
- 自动化执行从数据预处理到模型评估的完整训练闭环
- 支持SFT监督微调：用户自行准备训练数据，AI负责格式检查、配置确认和提交训练任务
- 智能选择训练策略：根据初始模型效果自动决定采用「先RFT再GRPO」或「直接GRPO」策略
- 标准化训练流程：严格遵循火山方舟ark-sdk最佳实践，确保训练任务成功率
- 关键节点用户确认：在重要决策环节寻求用户确认，避免误操作
前置条件
在执行训练流程前，根据训练类型检查不同文件：
- SFT训练：必须有用户自行准备的训练数据集文件（JSONL格式），验证集可选。
- RFT/GRPO/RL训练：必须有训练数据集文件、rollout函数代码文件、grader函数代码文件。
- 若用户提供的数据包含图片、视频、Function Calling或thinking字段，必须加载对应格式指南检查。 若缺失对应训练类型的必需文件，流程将终止并提示用户补充。
What ships with it
24 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- LICENSE 11 KB
- references/ark-sdk guide.md 9.0 KB
- references/RL guide.md 23 KB
- references/templates/job_grpo_lora.py 4.5 KB runs code
- references/templates/job_grpo_lora.yaml 4.0 KB
- references/templates/job_sft_lora.py 3.0 KB runs code
- references/templates/job_sft_lora.yaml 2.8 KB
- references/templates/README.md 2.8 KB
- references/模型精调数据集格式指南/CPT.md 2.6 KB
- references/模型精调数据集格式指南/DPO.md 19 KB
- references/模型精调数据集格式指南/Function Calling 样本要求.md 8.7 KB
- references/模型精调数据集格式指南/RL.md 6.0 KB
- references/模型精调数据集格式指南/SFT.md 17 KB
- references/模型精调数据集格式指南/图片文件要求.md 17 KB
- references/模型精调数据集格式指南/多轮reasoning_content的样本文件拆分.md 5.4 KB
- references/模型精调数据集格式指南/对视频内容进行抽帧处理.md 12 KB
- references/模型精调数据集格式指南/数据集Thinking字段处理工具.md 503 B
- references/模型精调数据集格式指南/视频文件要求.md 1.3 KB
- scripts/ark_trainer_helper.py 17 KB runs code
- scripts/modules/__init__.py 607 B runs code
- scripts/modules/endpoint.py 6.0 KB runs code
- scripts/modules/job.py 10 KB runs code
- scripts/modules/model.py 10.0 KB runs code
- scripts/modules/train.py 32 KB runs code
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 7d ago First seen · 736 lines · 107 tokens per session scan B 561703709cb0
byted-ark-trainer is a skill published in the GitHub repository bytedance/agentkit-samples (450 stars, last pushed 4d ago), licensed Apache-2.0. It adds 107 tokens to every session and 16,852 once invoked, about $0.0005 per session on Opus 5. A static security scan graded it B with 1 finding (asks for root). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
glycobiology
Glycosylation site prediction and glycobiology analysis. N-glycosylation motif finding, O-glycosylation hotspot prediction, glycan structure resources. Lightweight, pure Python. For protein function queries use uniprot-database; for structure analysis use alphafold-database.
shap
Model interpretability and explainability using SHAP (SHapley Additive exPlanations). Use this skill when explaining machine learning model predictions, computing feature importance, generating SHAP plots (waterfall, beeswarm, bar, scatter, force, heatmap), debugging models, analyzing model bias or fairness, comparing…
llm-redteam-overview
LLM red team category — full AATMF v3 tactic coverage (T01–T15). Routing skill: read this first to identify which tactic applies, then load the matching sub-skill. Maps to MITRE ATLAS where overlap exists.
vector-and-embedding-weaknesses
Hunt vector / embedding weaknesses (OWASP LLM08:2025) — adversarial inputs against the RAG / similarity layer that cause cross-tenant leak, embedding-inversion privacy loss, semantic confusion, and retriever-driven prompt injection.
groq-inference
Ultra-fast LLM inference on custom LPU hardware. OpenAI-compatible API at api.groq.com. Lowest latency in the industry (500-1000+ tok/s). Supports chat completions, vision, audio (Whisper STT + TTS), tool calling, JSON mode, and streaming. Free tier available. Inference only — no training.
aeon
This skill should be used for time series machine learning tasks including classification, regression, clustering, forecasting, anomaly detection, segmentation, and similarity search. Use when working with temporal data, sequential patterns, or time-indexed observations requiring specialized algorithms beyond standard…