fine-tuning-guide
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The
score_resultsfunction inSKILL.mdinterpolates untrusted model outputs into a prompt for LLM-based evaluation. This presents a potential surface for indirect prompt injection where malicious data in the test set could attempt to influence the scoring model. - Ingestion points: The
resultsparameter in thescore_resultsfunction inSKILL.md. - Boundary markers: Absent; data is directly interpolated into the string template.
- Capability inventory: Limited to model completion calls via the
client.chat.completions.createmethod; no system commands or external network access is triggered by this data. - Sanitization: No sanitization or escaping of the input variables is performed.
- [EXTERNAL_DOWNLOADS]: The skill references the
meta-llama/Llama-2-7b-hfmodel which is retrieved from Hugging Face using thetransformerslibrary. This is consistent with the skill's primary educational purpose and utilizes a well-known, reputable service.
Audit Metadata