dspy-better-together

Pass

Audited by Gen Agent Trust Hub on May 13, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides legitimate instructions for model optimization using the well-known DSPy library. Analysis of the provided scripts and instructions revealed no signs of prompt injection, data exfiltration, or unauthorized command execution.
  • [INDIRECT_PROMPT_INJECTION]: The skill demonstrates processing external datasets for model training and validation. While this creates a potential surface for indirect prompt injection if untrusted data contains malicious instructions, this is a standard requirement for the skill's intended purpose and is considered low risk within a development environment.
  • Ingestion points: External training and validation data loaded from JSON files (intents.json, reviews.json) or local variables.
  • Boundary markers: None explicitly defined in the provided code examples.
  • Capability inventory: Performs model fine-tuning via API calls and saves optimized programs to the local filesystem using compiled.save().
  • Sanitization: Standard data parsing and loading are employed; no specific input sanitization for instructions is shown in the snippets.
Audit Metadata
Risk Level
SAFE
Analyzed
May 13, 2026, 06:46 PM