Open-source command center for governed AI coding agents with built-in onboarding, hard gates, MCP, and shareable run receipts.
React Native SDK for Mushi Mushi — shake-to-report, bottom sheet widget, full context capture
ThumbGate self-improving agent governance: thumbs-up/down turns every mistake into a prevention rule and blocks repeat patterns. 36 pre-action checks, budget enforcement, and self-protection for Claude Code, Cursor, Codex, Gemini CLI, and Amp.
A command-line interface for Phoenix
React bindings for Mushi Mushi — Provider, hooks, and ErrorBoundary integration
Reference core for the Safe Automation Control Plane (SACP): policy engine, rule-only fallback, decision-output schema validation, decision state machine, circuit breaker and token-bucket rate limiter. Zero runtime dependencies, ports & adapters.
Local AI coding agent observability dashboard for Claude Code, Codex, OpenCode, OpenClaw, and Qoder. Track token usage, estimated cost, tool calls, subagent trees, and replay full sessions from local JSONL/SQLite files — no cloud, no telemetry.
OpenCode plugin for LLM prompt caching, token savings, and KV cache reuse across AI agent sessions
Vue 3 SDK for Mushi Mushi — plugin, composables, error handler integration
Svelte SDK for Mushi Mushi — context-based init, SvelteKit error hook integration
Capacitor plugin for Mushi Mushi — shake-to-report, screenshot capture, offline queue, and bottom-sheet widget for iOS/Android Capacitor apps.
Angular SDK for Mushi Mushi — standalone provideMushi(), injectable service, global ErrorHandler
Server-side Node SDK — framework middleware (Express / Fastify / Hono), unhandled-rejection hook, OTel trace propagation, bidirectional Sentry correlation. The 'glue' SDK for server-originated reports.