Security
Sensitive evaluations need visible controls, not vague trust copy.
Access
01
Workspace scoped access, SSO support, role-based permissions, and audit trails.
Secrets
02
Customer LLM API keys are encrypted at rest and are never exposed to the browser.
Data
03
Evaluation data is used to provide the product. Customer data is not used for model training.
Evidence
04
Assessment runs, reports, and changes stay traceable for security and compliance review.
Security posture follows the evidence posture.
Peeld handles prompts, endpoints, API keys, run results, and report exports as operational assurance data. The system is designed so teams can see who changed what, which target was tested, and what evidence was produced.