Claude Code usage
Fable5 2 Opus5 5 Orchestrator
Keep Claude Fable 5 in the chair all day without draining your usage limit — token-frugal multi-agent orchestration plugin for Claude Code: tier routing (Sonnet 5 does the work), Requirements Ledger, enforced guard hooks, automatic teammate reaping
Install
/plugin marketplace add Rylaa/fable5.2-opus5.5-orchestratorFound in the project README. Check it for setup steps after installing.
git clone https://github.com/Rylaa/fable5.2-opus5.5-orchestrator.gitFrom the README
Fable Orchestrator Run Claude Fable 5.2 all day — without watching the usage meter. Fable 5.2 is the best chair a Claude Code session can have, and the most expensive one. This plugin keeps it on planning and decisions and sends the work to named workers: the volume to Sonnet 5.5, the hard slices (architecture, irreversible migrations, security review) straight to Opus 5.5. No tier ever runs above high effort. The division of labor - Escalation is one-way. When high is not enough, the work moves up a tier, not up in effort. Sonnet's "uncertain" goes to Opus, never back to the chair. - A declined task is re-run unchanged on another tier, never reworded. If that tier declines too, the chair stops and tells you. What the plugin does 1 · The chair profile. Every session starts with a ~3.7k-char core: delegate by default, hand off through ./.workflow/scratch/, name the workers, set model and effort per spawn. The detail is in the orchestrator:playbook skill, loaded before the first delegation. Worker reports are capped at 40 lines; longer material goes to disk. 2 · The solo guard. The chair's own Edit / Write / MultiEdit / NotebookEdit calls are counted. The 3rd one in a session that has spawned no worker is denied once, then the guard stays quiet. One spawned worker disarms it. A fork does not count as a worker. 3 · The naming gate. The first unnamed spawn with a 1500+ char prompt is denied once. Named workers run in tmux panes you can watch; unnamed ones are invisible until they return. 4 · The per-prompt reminder. One line rides every prompt: *You are the CHAIR. Delegate by default.* Workers (teammates) get none of this — the rules are for the chair. When a session ends, its teammate panes are killed. Two chairs Fable is primary. When its limit runs out, move…
Details
- Repository
- Rylaa/fable5.2-opus5.5-orchestrator
- Language
- Python
- License
- MIT
- Last updated
- 11 November 2026
- Topics
- ai-agentsanthropicclaudeclaude-codeclaude-code-pluginclaude-fablecontext-managementfablefable-5hooksmulti-agentorchestration
FAQ
What is Fable5 2 Opus5 5 Orchestrator?
Keep Claude Fable 5 in the chair all day without draining your usage limit — token-frugal multi-agent orchestration plugin for Claude Code: tier routing (Sonnet 5 does the work), Requirements Ledger, enforced guard hooks, automatic teammate reaping It is a usage for Claude Code, from Rylaa/fable5.2-opus5.5-orchestrator on GitHub.
How do I install Fable5 2 Opus5 5 Orchestrator?
Run /plugin marketplace add Rylaa/fable5.2-opus5.5-orchestrator inside Claude Code, then follow the setup in the README.
Is Fable5 2 Opus5 5 Orchestrator popular?
Rylaa/fable5.2-opus5.5-orchestrator has 79 GitHub stars and 26 forks. It was last updated on 11 November 2026.
More usage & cost trackers
Claude Usage Widget
81bozdemir/claude-usage-widget
Cross-platform desktop widget + CLI for real-time Claude Code (and opt-in OpenAI Codex) usage — session/weekly limits, cost, forecasts, burn alerts, 11 themes. One pip install (Linux/macOS/Windows).
Clawd Usage Winwidget
88Enhed/clawd-usage-winwidget
Claude usage limits in a tiny Windows widget + tray icon — with pixel Clawd who reacts to your Claude Code sessions, looks after your PC and chats with you, from the widget or your phone. PowerShell, no dependencies.
Claude Code Meter
92gxjansen/claude-code-meter
macOS desktop widget showing Claude Code usage limits (5-hour and 7-day windows) in real time. Built for Ubersicht.
Agent Monitor
86donvito/agent-monitor
Visualize your Codex and Claude Code subagents. Advanced traces view, tokens breakdown, and cost usage is also supported.
Claude Meter
88abhishekray07/claude-meter
Tracking Claude Code Usage Limits
Sidekick Agent Hub
89cesarandreslopez/sidekick-agent-hub
See what your AI coding agent is doing. Multi-provider assistant & session monitor for VS Code and the terminal — inline completions, code transforms, and a full TUI dashboard tracking tokens, cost, and context across Claude Code, OpenCode, and Codex. Switch between multiple Claude Code and Codex accounts.
Browse the most popular Claude Code mods, all usage & cost trackers, or the skills and MCP server directories.