Strictly evaluates AI agent outputs against acceptance policies for quality and compliance.
Copy the install command and let the AI configure it · recommended for beginners
Please install the "ai.seaotter/otterscore" MCP server from askskill: Run: claude mcp add --transport http 'ai-seaotter-otterscore' 'https://mcp.seaotter.ai/mcp'
Evaluate this AI customer support reply against the following acceptance policy: it must accurately explain the refund policy, remain polite, avoid invented terms, and provide next steps. Score each criterion, identify failures, and give a final pass/fail verdict. Reply: {{agent_output}}A strict evaluation with criterion scores, issue notes, and a final pass/fail decision.
Act as a hostile-by-default reviewer and check this AI-generated code explanation and patch against the acceptance criteria: it must meet requirements, preserve existing interfaces, cover edge cases, and avoid obvious security risks. Output detailed deduction reasons and improvement suggestions. Content: {{agent_output}}A strict review report for the code output with risks, scoring rationale, and fixes.
Score this AI research summary using the following acceptance policy: it must stay faithful to the source conclusions, note uncertainty, avoid overstating evidence, and be clear for executives. Return a total score, category assessments, key defects, and whether you recommend using it. Summary: {{agent_output}}A validation result for research or product teams that clearly states usability and defects.
Evaluate your judgment in AI-assisted work and improve decision quality.
Assess EU AI Act compliance locally, obligations, deadlines, and documentation gaps.
Test API compatibility for AI agents with scores, grades, and recommendations.
Generate test scenarios and measure how well agents follow skills and rules.
Review contract code with multi-agent AI and get verdict-driven quality insights.
Run multi-agent debates and produce consensus recommendations with ranked options.