Frugal Fable Planner-Worker Orchestrator
Keeps a premium Claude model as planner and delegates coding work to cheaper sub-agents with checkpoints. Stretches weekly usage limits without losing judgment.
Before & after
Same prompt, same agent — the only difference is the skill.
To manage usage limits on a large build, you could break the work into smaller chunks and be strategic about which parts need the most capable model. Some ideas: - Use a more capable model for planning and architecture - Use a cheaper model for routine implementation - Break the work into phases - Review outputs before continuing The most token-intensive work is often the back-and-forth during implementation, so having a clear plan upfront reduces wasted effort. How large is the build, and what does it involve?
Right instinct — the burn is in implementation, not planning. This wraps
your build in a planner-first flow that keeps the premium model OUT of
the code entirely.
THE FLOW
1. PLANNER (premium model) reads the PRD, asks clarifying questions,
produces a spec-driven multi-phase architecture. This is where
judgment matters and it's cheap — planning is a fraction of a
build's tokens.
2. Each phase hands off to a CHEAPER sub-agent, matched to the work:
Research / exploration → Haiku (cheapest, fine for reading)
Scaffolding / boilerplate → Sonnet
Heavier implementation → Opus 4.8
3. AUTOMATIC CHECKPOINTS + self-review between phases, so a cheap
worker going off-spec is caught before it compounds.
THE RULE THAT SAVES THE TOKENS
The premium model NEVER touches implementation. It plans, it reviews at
checkpoints, and it delegates. Implementation — the token-heavy
back-and-forth you're worried about — runs on cheaper models. That's the
whole savings mechanism: you're paying premium rates for judgment, cheap
rates for typing.
WHY CHECKPOINTS MATTER HERE
A cheap worker with no oversight drifts, and fixing drift costs more than
it saved. The checkpoints let the planner catch a phase that went sideways
and correct course — so you get cheap-model economics with premium-model
judgment on the decisions that matter.
WHAT YOU GET
A long build that would have burned your weekly limit runs mostly on
sub-agents, with the premium model spending its tokens on the plan and the
checkpoints — the two places its judgment actually changes the outcome.
Point me at the PRD and I'll run the planner pass first: clarifying
questions, then the phased architecture with a model assigned per phase.About this skill
name: premium-planner-cheap-worker-orchestrator description: Keeps a premium Claude model as planner and delegates coding work to cheaper sub-agents with checkpoints. Stretches weekly usage limits without losing judgment. Use when working on frugal fable planner-worker orchestrator.
Frugal Fable Planner-Worker Orchestrator
Wraps any long-running Claude Code task in a planner-first flow. The premium model reads the PRD, asks clarifying questions, and produces a spec-driven multi-phase architecture. Each phase is handed to a cheaper sub-agent (Haiku for research, Sonnet for scaffolding, Opus 4.8 for heavier code) with automatic checkpoints and self-review passes. Prevents the premium model from touching implementation and keeps weekly usage burn low. Use when a long build would otherwise exhaust your weekly premium tier.
What you get
- Spec-driven implementation plan
- Sub-agent delegation map
- Per-phase checkpoints and review passes
- Usage-limit-aware execution loop
- Weekly burn report
- Manual approval gate before merge
Customize your output
- Premium tier model: Opus, Fable 5
- Cheap tier model: Haiku, Sonnet, Opus 4.8
- Max percent of weekly budget: 25, 50, 75
- Number of phases: auto, 4, 6, 8, 12
- Verification depth: light, strict
- Approval gates: auto, per phase, manual
Example output
12-phase Swift medical-doc app plan. Opus 4.8 handles Phase 1 OCR module, Sonnet handles UI scaffolding, Haiku handles research. Planner spends 8 percent of weekly usage over 3 hours instead of the usual 100 percent burn.
Best for
Heavy Claude Code and Codex users, agencies running production dev work on subscription tiers, technical founders on usage-capped plans.
Note: Depends on Claude usage-limit behavior that may change. Skill exists in the original author's marketplace. High confidence.
SKILL.md preview
---
name: premium-planner-cheap-worker-orchestrator
description: Use this skill when a long Claude Code build would otherwise exhaust the weekly premium usage tier and the user wants a premium planner to delegate implementation to cheaper sub-agents.
version: 1.0.0
category: AI Systems
author: AgentVolt
license: proprietary
tags:
- ai-systems
- p4
---
# Frugal Fable Planner-Worker Orchestrator
Wraps a long-running build in a planner-first flow: a premium model plans and reviews while cheaper sub-agents handle the actual implementation, stretching weekly usage limits without losing architectural judgment.
## When to use
… (sign up to view the full skill)