engine-sync

Multi-agent framework for Claude Code, Codex & OpenCode — coding agents you can watch, message and hold to rules

The model writes its own handoff through one tool and the plugin installs it at Claude Code's compaction, with facts recorded from typed events and every later message kept verbatim.

A /cache command that switches the prompt-cache TTL (5m, 1h or automatic) for one chat and its sub-agents, live, with no model turn.

Tune reasoning levels — automatic main-chat tuning; set_effort, effort_up, effort_down and effort_auto choices; /effort pins; [effort: low|medium|high|xhigh|max] sub-agent launches with bounded failure escalation.

Claude Code plugin for per-agent tool visibility — hide and refuse subagents, skills, MCP and built-in tools per loop

Claude Code buddy plugin: Quack the ASCII duck waddles above your prompt and talks back — or bring back your /buddy

Claude Code plugin for tracking token usage and context growth — every tool call and sub-agent in local SQLite

Keeps the main chat's prompt cache warm while it waits on background agents, shells, Monitors and workflows: one invisible replay before expiry, admitted only while the forecast spend stays under the rewrite it saves.

Fetches, converts, caches and searches documents through the harvester CLI and MCP tools.