Skip to content

Retinue

Self-hosted coordination for AI workers: per-task graphs, device/model timelines, progress receipts, and source-aware usage. MIT.

Install

$git clone https://github.com/jklthinking/retinue

Install command taken from the project's README or registry entry. Always review code before running it.

Retinue (众卿) — self-hosted task board for humans and AI agents

English · 简体中文 · English demo · 中文演示

The dashboard now switches between English and 简体中文. The language preference is saved locally and preserves the selected task/page. User-authored task content is retained. See the English PRD, English screenshots and release notes, and English sharing draft.

Retinue (Chinese name 众卿, "the assembled ministers") is a self-hosted task board and orchestration hub where a mixed team of people and AI agents work the same cards. Every piece of work is one durable baton: a task card with a single holder, an acceptance check, and an append-only receipt chain. You run the board, agents claim work over MCP or HTTP, they write back, and you accept the card or send it back. There is no hosted control plane, no vendor account requirement, and no telemetry. The server-backed hub uses operator-managed local accounts and scoped credentials.

2026-10-07: model quota on Home

Version 0.3.0a4 adds task conversation histories behind an opt-in flag. Release notes. Manual quota refresh remains available.

Opt-in node quota collection and bilingual Home cards for Claude, Codex, Grok, Cursor and Kimi subscriptions, plus Moonshot API balance. See used percentages, reset time, report age and labelled last-successful data. Setup and API semantics · Update notes · Conversation timeline design. Existing servers require the explicit schema-26 migration; collectors stay off until enabled locally. The online demos show synthetic data only.

2026-10-03 English edition: visible collaboration

Version 0.3.0a1 brings the collaboration process into each task. See who delegated to whom, what each device/model worker reported, what it is waiting for, and which artifact supports the next handoff.

View What it explains
Home The original task flow, dispatch coordination, and session flow views
Per-task graph Explicit delegation, dependencies, baton movements, and review returns
Device/model timeline Recorded runs with identity snapshots and source labels
Module contributions Explicit functional assignments, reported work, remaining items, and artifact references
Node details and context Instructions, acceptance criteria, waiting owners, progress, evidence versions, and next-step intent
Operations and data quality Separate sources, reported usage coverage, receipt freshness, and retained history

Read-only cc-connect observers match native sessions to Claude records and deduplicate usage by stable deliveries. A worker is an execution identity; session transports and observers are listed separately. The operations view has one primary implementation, while the existing task and session modes remain available.

Similar plugs

The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.

Java 面试 & 后端通用面试指南,覆盖计算机基础、数据库、分布式、高并发、系统设计与 AI 应用开发

Build Agentic workflows, RAG pipelines, with rich AI model and tool support on one collaborative workspace. Deploy on cloud, VPC, or self-hosted, so teams move from prototype…

A cross-platform desktop All-in-One assistant for Claude Code, Codex, OpenCode, OpenClaw, Grok Build & Hermes Agent. Only official website: ccswitch.io