eVi feature guides

Deep-dive docs for every feature area — overview, how it works, setup, usage (CLI/REPL/Web), and worked examples. For the one-page catalog see ../features.md; for config keys see ../configuration.md; for surface coverage see ../cli-parity.md.

Guides

• Agents & Orchestration — subagents, delegate, parallel research, workflows, dispatch, federation
• Ultracode — exhaustive multi-agent pipeline for one hard task (evi ultracode / /ultra)
• Content Guardrails — regex + LLM-judge + offline classifier rules
• Plugins & Marketplace — bundled commands/skills/hooks/MCP/agents + plugin search/install
• Skills — markdown instruction packets the model loads on demand (invoke_skill)
• Slash commands — built-in /cmd controls + user/plugin command templates
• Evals & LLM-as-judge — evi eval suites, judge rubrics, scheduled evals
• Voice — TTS engines (system/coqui/f5/piper), STT, AutoSpeaker/Listener
• Sessions, Resume, Handoff, Checkpoints
• Hooks — tool + lifecycle events, command or HTTP
• Structured Outputs & Batch — JSON-Schema output + evi batch
• Permissions & Sandbox — modes, rules, trusted dirs/domains, OS sandbox
• MCP — client (mcp.json) + server (evi mcp serve)
• Recipes, Routines, Scheduled tasks, Channels
• Web & Desktop — settings, multi-user, deep links, updater, working indicator
• Observability — OpenTelemetry, evi stats, crash reports
• Memory & Context — memory tags, /context, compaction