By name: promptfoo description: Test and evaluate LLM prompts systematically. allowed-tools: Read, Write, Glob, Grep, Task
Promptfoo
Category: Performance EU AI Act Reference: Art. 15 Source: 3rd-party (Promptfoo)
When to Use This Skill
Use this skill when:
- Promptfoo tasks - Working on test and evaluate llm prompts systematically
- Compliance assessment - Evaluating AI systems against Art. 15 requirements
- Risk mitigation - Implementing controls for performance risks
Overview
Test and evaluate LLM prompts systematically. Run evaluations across multiple models, detect regressions, and benchmark prompt performance.
Key Requirements
Regulatory Context (Art. 15)
This tool supports compliance with the following EU AI Act requirements:
- Assessment - Evaluate current state against Art. 15 requirements
- Implementation - Deploy appropriate controls and measures
- Documentation - Record findings, decisions, and evidence
- Monitoring - Establish ongoing monitoring and review processes
Assessment Checklist
- Identify applicable Art. 15 requirements
- Assess current compliance status
- Document gaps and findings
- Develop remediation plan
- Implement required controls
- Verify effectiveness of measures
- Maintain ongoing documentation
Integration Notes
Source Type: 3rd-party (Promptfoo) Category: Performance