Agent capability
eval-driven-dev
Improve AI application with evaluation-driven development. Define eval criteria, instrument the application, build golden datasets, observe and evaluate application runs, analyze results, and produce a concrete action plan for improvements. ALWAYS USE THIS SKILL when the user asks to set up QA, add tests, add evals, evaluate, benchmark, fix wrong behaviors, improve quality, or do quality assurance for any Python project that calls an LLM model.
Runtime provisioning
Provision this capability
The provisioning flow resolves the versioned manifest, verifies every SHA-256 digest, and applies the package atomically.
Shared Agent Skills
User scope: ~/.agents/skills/eval-driven-dev
curl -fsSLo /tmp/mm-skills-install.sh __MM_RUNTIME_ORIGIN__/install.sh && bash /tmp/mm-skills-install.sh eval-driven-dev --client agents --scope user --registry __MM_RUNTIME_ORIGIN__
Codex
User scope: ~/.agents/skills/eval-driven-dev
curl -fsSLo /tmp/mm-skills-install.sh __MM_RUNTIME_ORIGIN__/install.sh && bash /tmp/mm-skills-install.sh eval-driven-dev --client codex --scope user --registry __MM_RUNTIME_ORIGIN__
OpenCode
User scope: ~/.config/opencode/skills/eval-driven-dev
curl -fsSLo /tmp/mm-skills-install.sh __MM_RUNTIME_ORIGIN__/install.sh && bash /tmp/mm-skills-install.sh eval-driven-dev --client opencode --scope user --registry __MM_RUNTIME_ORIGIN__
Antigravity
User scope: ~/.gemini/config/skills/eval-driven-dev
curl -fsSLo /tmp/mm-skills-install.sh __MM_RUNTIME_ORIGIN__/install.sh && bash /tmp/mm-skills-install.sh eval-driven-dev --client antigravity --scope user --registry __MM_RUNTIME_ORIGIN__
Windows PowerShell
$installer = Join-Path $env:TEMP 'mm-skills-install.ps1'; Invoke-WebRequest '__MM_RUNTIME_ORIGIN__/install.ps1' -OutFile $installer; & $installer -Skill 'eval-driven-dev' -Client 'agents' -Scope 'user' -Registry '__MM_RUNTIME_ORIGIN__'
Artifact integrity