huggingface-llm-trainer
This skill should be used when users want to train or fine-tune language models using TRL (Transformer Reinforcement Learning) on Hugging Face Jobs infrastructure. Covers SFT, DPO, GRPO and reward modeling training methods, plus GGUF conversion for local deployment. Includes guidance on the TRL Jobs package, UV scripts with PEP 723 format, dataset preparation and validation, hardware selection, cost estimation, Trackio monitoring, Hub authentication, and model persistence. Should be invoked for tasks involving cloud GPU training, GGUF conversion, or when users mention training on Hugging Face Jobs without local GPU setup.
72
88%
Does it follow best practices?
Run evals on this skill
Adds up to 20 points to the overall score
View guide
Critical
Do not install without reviewing
Security
2 findings: 1 critical severity, 1 high severity. Installing this skill is not recommended: please review these findings carefully if you do intend to do so.
E006: Malicious code pattern detected in skill scripts.
What this means
Detected high-risk code patterns in the skill content — including its prompts, tool definitions, and resources — such as data exfiltration, backdoors, remote code execution, credential theft, system compromise, supply chain attacks, and obfuscation techniques.
Why it was flagged
The codebase includes explicit calls that enable execution of remote/untrusted code and installs dependencies directly from a freshly cloned repository, creating clear remote code execution / supply-chain risk (potential backdoor vector).
W007: Insecure credential handling detected in skill instructions.
What this means
The skill handles credentials insecurely by requiring the agent to include secret values verbatim in its generated output. This exposes credentials in the agent’s context and conversation history, creating a risk of data exfiltration.
Why it was flagged
The skill repeatedly instructs embedding the user's HF_TOKEN value into job configs/CLI (e.g., secrets={"HF_TOKEN": "$HF_TOKEN"} and explains $HF_TOKEN is the actual token), which requires the LLM to handle or output the secret verbatim and therefore is high risk for secret exfiltration.
Low
Low-risk findings.
1 low severity finding. Worth noting, but not necessarily harmful.
W012: Unverifiable external dependency detected (runtime URL that controls agent).
What this means
The skill fetches instructions or code from an external URL at runtime, and the fetched content directly controls the agent’s prompts or executes code. This dynamic dependency allows the external source to modify the agent’s behavior without any changes to the skill itself.
Why it was flagged
The skill instructs fetching and executing remote scripts at runtime—for example cloning and running code from https://github.com/ggerganov/llama.cpp.git in scripts/convert_to_gguf.py and invoking the remote dataset inspector at https://huggingface.co/datasets/mcp-tools/skills/raw/main/dataset_inspector.py via hf_jobs, which are runtime downloads that execute remote code and are relied on by the skill.
- Repository
- openai/plugins
- Commit
11c74d6
- Audited
- Security analysis
Is this your skill?
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.