Add Docker skill eval workbench

[bucurdavid]

Summary

• Replace the old benchmark/optimizer stack with the Docker workbench CLI for cases, suites, graders, traces, trials, and reference-solution verification.
• Add example workbench suites and package the canonical skill-optimizer skill.
• Add plugin distribution for Claude Code, OpenCode, Codex, Cursor, and Gemini.

Test Plan

• npm test
• npm run typecheck
• npm pack --dry-run --json package verifier

[Copilot]

Copilot wasn't able to review this pull request because it exceeds the maximum number of lines (20,000). Try reducing the number of changed lines and requesting a review from Copilot again.