[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-92568":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":10,"language":10,"languages":10,"totalLinesOfCode":10,"stars":11,"forks":12,"watchers":13,"openIssues":14,"contributorsCount":14,"subscribersCount":14,"size":14,"stars1d":14,"stars7d":14,"stars30d":15,"stars90d":14,"forks30d":14,"starsTrendScore":14,"compositeScore":16,"rankGlobal":10,"rankLanguage":10,"license":17,"archived":18,"fork":18,"defaultBranch":19,"hasWiki":20,"hasPages":18,"topics":21,"createdAt":10,"pushedAt":10,"updatedAt":22,"readmeContent":23,"aiSummary":24,"trendingCount":14,"starSnapshotCount":14,"syncStatus":13,"lastSyncTime":25,"discoverSource":26},92568,"fable-advisor","DannyMac180\u002Ffable-advisor","DannyMac180","Claude Fable as an orchestrator for Opus, GPT and Sonnet","",null,378,28,2,0,256,54.39,"MIT License",false,"main",true,[],"2026-07-22 04:02:06","# Fable Advisor\n\n**The smartest model runs the show. Cheaper models do the typing.**\n\nClaude Code lets every subagent run on a different model — and lets the session itself run on a different model than its subagents. This plugin exploits that with the **architect pattern**: your session runs on **Fable 5**, Anthropic's most capable model, acting as a full-time architect. It owns requirements, decomposition, specs, and verification — and routes every implementation task to the cheapest adequate lane:\n\n| Lane | Producer | Invocation | Route here when |\n|---|---|---|---|\n| Routine | **Grok 4.5** | `grok-implementer` agent (default) | The spec fully determines the outcome — Grok does the typing via the [Grok CLI](https:\u002F\u002Fx.ai\u002Fcli) |\n| Cross-vendor | GPT-5.5 | `codex-implementer` agent | Correctness-critical, or you want a second independent implementation to compare |\n| Judgment | Fable 5 | `fable-advisor` agent | Commitment boundaries — see below |\n\nTokens route by volume: the expensive model emits the fewest tokens (judgment and specs), cheap lanes emit the most (code). Implementation mechanics are ~90% of a session's tokens and Grok 4.5 handles them at near-parity — so this runs far cheaper than Fable-for-everything, and every implementation comes from a *different model family* than the architect that reviews it: cross-vendor review is built into the routing, not bolted on. For high-stakes work, race `grok-implementer` and `codex-implementer` on the same spec and let the architect pick the stronger diff.\n\nThe plugin ships the **orchestration skill** — the routing doctrine that teaches the session when to use each lane, the cost discipline that keeps the expensive model's own token volume minimal (emit judgment not volume, keep context lean, reason once then hand off), the five-part spec contract that makes context-free delegation safe, and the verification rules that keep cheap lanes honest.\n\n## Install\n\n```\nclaude plugin marketplace add DannyMac180\u002Ffable-advisor\nclaude plugin install fable-advisor\n```\n\nThen start your session as the architect:\n\n```\n\u002Fmodel fable\n```\n\n**Lite mode — one file, 30 seconds.** Don't want the full pattern? Copy [`agents\u002Ffable-advisor.md`](agents\u002Ffable-advisor.md) into `~\u002F.claude\u002Fagents\u002F` and keep your session on Sonnet. You get advisor consults at commitment boundaries without the orchestration layer (see \"Advisor-only mode\" below).\n\n## Requirements\n\n- **Claude Code ≥ 2.1.170** with a subscription that includes Fable 5 (Pro, Max, Team, or Enterprise — all current consumer plans qualify).\n- **No Fable access** (e.g. API-key billing)? Use `\u002Fmodel opus` for the session and change `model: fable` → `model: opus` in the advisor file. Same pattern, model tiers shift down one.\n- **Grok lane (the default implementer):** the `grok-implementer` agent needs the [xAI Grok CLI](https:\u002F\u002Fx.ai\u002Fcli) installed and authenticated (install from [x.ai\u002Fcli](https:\u002F\u002Fx.ai\u002Fcli), then `grok login`). It drives **Grok 4.5** headlessly (`grok --prompt-file … -m grok-4.5`). Without it the agent reports `STATUS: unavailable` — it never silently falls back to a Claude model.\n- **Codex lane (optional):** the `codex-implementer` agent needs the [OpenAI Codex CLI](https:\u002F\u002Fgithub.com\u002Fopenai\u002Fcodex) installed and authenticated (`npm i -g @openai\u002Fcodex`, then `codex login`). Without it the agent reports `STATUS: unavailable` — the other lanes are unaffected.\n- Heads-up: if a pinned Claude model isn't available on your account, Claude Code silently falls back to your session model — the pattern degrades quietly rather than erroring. If results feel unremarkable, check your plan. (This quiet fallback applies only to Claude model pins — the grok and codex lanes always fail loudly with a structured error.)\n\nModel resolution order in Claude Code: `CLAUDE_CODE_SUBAGENT_MODEL` env var → per-invocation `model` parameter → agent frontmatter → session model.\n\n## Use it\n\nWith the session on Fable, just ask for work — the orchestration skill routes it:\n\n```\nAdd rate limiting to our public API. Design it, delegate the\nimplementation, and verify the evidence before you call it done.\n```\n\nThe architect writes the spec, picks the lane (rate limiting touches concurrency — a good case for racing `grok-implementer` against `codex-implementer` and picking the stronger diff), reads the diff and verification evidence when the report comes back, and only then reports done.\n\nTo make the doctrine always-on, add one line to your project's `CLAUDE.md`:\n\n```\nYou are the architect running the most expensive model — minimize your\nown token volume. Delegate all implementation through the orchestration\nskill's routing table (never type code yourself), delegate broad codebase\nexploration to cheap read-only agents, and verify evidence before\naccepting any lane's report.\n```\n\n## Commitment boundaries\n\nEven the architect gets a second opinion. The `fable-advisor` agent is a read-only skeptic — consulted before architecture decisions, migrations, API designs, and whenever a problem has resisted two attempts. It reads your actual code and returns a verdict in under 300 words. It never implements. Running it from a Fable session still pays: it sees the code fresh, without your conversation's accumulated assumptions.\n\n## Advisor-only mode (the original pattern)\n\nThe inverse arrangement, for when you'd rather keep the session cheap: run the session on Sonnet and consult `fable-advisor` only at commitment boundaries.\n\n```\nMigrate our checkout sessions from Postgres to Redis — plan it,\nconsult your advisor before committing, then implement.\n```\n\nA typical consult costs cents. To make it automatic, add to your project's `CLAUDE.md`:\n\n```\nBefore committing to any architecture decision, migration, or refactor\ntouching 3+ files, consult the fable-advisor agent and act on its verdict.\n```\n\n## FAQ\n\n**Is this Anthropic's \"advisor tool\"?** No — that's a server-side API feature. These are plain Claude Code subagents plus a skill: readable, editable, no beta flags.\n\n**Does this work on claude.ai?** No — subagent model routing is Claude Code only (CLI, desktop, VS Code, web).\n\n**Why not just run everything on Fable?** You can. It's excellent. It's also the most expensive lane per token, and most of a session's tokens are implementation mechanics that the cheap lanes handle at near-parity. Spend the premium where judgment lives.\n\n**Upgrading from v2?** v3 replaces the Sonnet\u002FOpus `implementer` agent with `grok-implementer` — Grok 4.5 via the [Grok CLI](https:\u002F\u002Fx.ai\u002Fcli) is now the default typing lane. The `fable-advisor` and `codex-implementer` agents are unchanged, and advisor-only mode works exactly as before. If you preferred the Claude implementer, grab [`implementer.md` from the v2.1.0 tag](https:\u002F\u002Fgithub.com\u002FDannyMac180\u002Ffable-advisor\u002Fblob\u002F3c1846c\u002Fagents\u002Fimplementer.md).\n\n**Why Grok and GPT-5.5 lanes in a Claude plugin?** Vendor diversity. Models from one family share blind spots; an independent implementation from a different lineage catches what same-family review misses — and with Claude as the architect, *every* diff now gets cross-vendor review for free. The architect stays Claude — the lanes are producers, not judges.\n\n## License\n\nMIT\n","Fable Advisor 是一个基于 Claude Code 的模型编排插件，实现多模型协同的智能任务分解与执行。它采用“架构师模式”，由高能力模型（如 Fable 5）担任全局架构师，负责需求理解、任务拆解、规格定义与结果验证；而具体实现则路由至成本更低的专用模型（如 Grok 4.5、GPT-5.5），按任务特性自动分派至 Routine、Cross-vendor 或 Judgment 路径。项目通过令牌量感知路由、五段式规格契约和跨厂商校验机制，在保障质量前提下显著降低推理成本。适用于需要高可信度、多模型交叉验证的复杂软件开发、系统设计与技术方案评审等场景。","2026-07-09 02:30:21","CREATED_QUERY"]