Benchmark analysis agent that surfaces patterns in eval results that aggregate stats might hide. Identifies failure clusters, reliability issues, and regression risks across iterations. User says: "analyze the benchmark results" User says: "what patterns do you see in the eval failures?".
Architecture design specialist for Claude Code skills. Analyzes use cases, determines complexity tier (1-4), plans file structure, routing tables, and sub-skill decomposition. User says: "design the architecture for a new DevOps skill" User says: "what tier should my skill be?".
Blind comparison agent for A/B testing skill versions. Evaluates outputs from two skill versions without knowing which is "new" vs "old" to eliminate bias. User says: "compare these two skill versions" User says: "run a blind A/B test on the skill".
Multi-platform skill conversion specialist for Claude Code, OpenAI Codex, Gemini CLI, Google Antigravity, and Cursor. Analyzes skills for cross-platform compatibility, identifies Claude-specific features, suggests adaptation strategies, and assesses conversion risk. User says: "can this skill work on Codex?" User…
Eval execution agent that runs skills against eval prompts and captures outputs, timing data, and token usage. Operates in isolated context to prevent eval bleed. User says: "run this skill against the eval prompts" User says: "execute eval set for my skill".
Eval grading agent that evaluates skill outputs against defined assertions. Checks each assertion, provides pass/fail with evidence, and calculates per-eval pass rates. User says: "grade the eval results" User says: "check if the outputs pass assertions".
Skill quality validation specialist. Runs programmatic and manual checks on Claude Code skills, generates health scores (0-100), and identifies issues by priority level. User says: "validate my skill" User says: "check skill quality".
SKILL.md content generation specialist. Writes high-quality frontmatter, descriptions, and instructions for Claude Code skills following the Agent Skills standard. User says: "write the SKILL.md for my tool" User says: "generate the skill content".