# Fable Opus

> Run fable-mode execution discipline on Claude Opus — one rung below Fable 5.1 on the escalation ladder. Routes the task to the @fable-orchestrator agent (Opus, Write-less), which stages the work, delegates artifact production to @fable-worker-sonnet / @fable-worker-haiku, and cold-checks deliverables with @fable-verifier. Trigger when the user explicitly asks for thorough/systematic/"deep work" handling on Opus ("fable on opus", "stage this on opus", "deep work mode, opus"). Escalate to fable-fable when the orchestrator reports the task exceeds Opus's ceiling. Do NOT use for ordinary single-pass tasks — and prefer fable-sonnet or fable-haiku when the task doesn't need peak reasoning.

- Skill: `mrtooher/fable-opus` (Agent Skill)
- Install (CLI): `npx skillmds@latest add mrtooher/fable-opus`
- Raw SKILL.md: https://api.skillmd.com/api/skills/mrtooher/fable-opus/raw
- Safety review: pending (external: skill-scanner PASS, skillspector PASS)
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: AI & ML
- Author: mrtooher (https://skillmd.com/u/mrtooher)
- Updated: 2026-09-10
- Page: https://skillmd.com/skills/mrtooher/fable-opus

---


# Fable Mode — Opus (v3, agent-routed)

v3 change: delegation is enforced structurally, not requested in prose. The
orchestrator is a real agent definition (`agents/fable-orchestrator.md`) with no
Write/Edit tool — it cannot do the work inline, so "spawn a worker" stops being
a suggestion the model can skip. (Change prompted by field report: prose-level
"you may spawn workers" almost always ran inline on the main thread.)

If a task has one obvious correct approach and fits in a single pass, skip this
loop and do it directly.

## How to run it

1. Confirm the fable agents are installed (`fable-orchestrator`,
   `fable-worker-sonnet`, `fable-worker-haiku`, `fable-verifier` appear in the
   available agent types). If they are not, fall back to the inline method:
   spawn a general-purpose Opus agent and pass it the Core Loop, domain
   patterns, and operational rules verbatim from the **fable-mode** skill,
   which is installed alongside this one. Do NOT try to read
   `agents/fable-orchestrator.md` — agent files exist only on surfaces that
   load an `agents/` directory (Claude Code), never on skill-only surfaces
   (Cowork), which is exactly where this fallback fires.
2. Spawn **@fable-orchestrator** via the Task tool (`subagent_type:
   "fable-orchestrator"`). Brief it with: the user's task, the output
   directory, relevant session context, and any user-set limits (warning
   threshold, worker cap, deadline).
3. Do not restate the Core Loop or operational rules in the briefing — the
   orchestrator's agent definition carries them. Brief the task, not the
   method.
4. When it returns, relay the result, every stage it marked unverified, and its
   recommendations (surfaced scope it did not build). If it reports the task
   beyond Opus's capability, offer a fable-fable rerun (Fable 5.1) rather than
   retrying Opus.
5. **Mandatory delivery gate:** before presenting the result to the user, invoke
   the **double-check** skill on the finished deliverable. Pass it the
   orchestrator's closing `FINAL-COLD-CHECK:` line verbatim — `yes` means
   fable-verifier already checked the finished deliverable and double-check
   runs the seam check only; `intermediates-only`, `no`, or a missing line
   means the full panel runs. Do not skip the gate; the user relies on it
   instead of re-checking the work themselves.

## Known limitation

Write-removal closes the front door, not the side door: the orchestrator keeps
Bash for running verification commands, and Bash can technically create files.
Its definition forbids that use; if audits show it writing through Bash, the
next tightening is removing Bash and routing even check-execution through
workers (cost: one extra hop per check).