distill-test-set
Pass
Audited by Gen Agent Trust Hub on Jun 19, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the official Coval CLI (
coval) and a local Python script (scripts/distill.py) to manage and process dataset files. These operations are performed locally and align with the skill's primary function. - [DATA_EXFILTRATION]: The skill interacts with the Coval API (
api.coval.dev) and OpenAI API (api.openai.com) for test set management and optional semantic clustering. These network operations are targeted at vendor-owned or well-known service endpoints and are used for their intended purposes. - [PROMPT_INJECTION]: The skill processes external data from CSV, JSON, and NDJSON files. Although this creates a surface for indirect prompt injection, the skill's logic is restricted to data distillation and sampling, and it does not interpolate the content into instructions that could override the agent's behavior.
Audit Metadata