🔓 unclaimed — this page was auto-generated from GitHub. Are you the creator?
Claim this page →
fable-mode
A Claude skill that activates Fable-style agentic behavior: explicit multi-stage planning, sub-agent delegation, and self-verification.
git clone https://github.com/mrtooher/fable-mode
fable-mode
A Claude skill that enforces staged execution discipline on large tasks: a written stage plan, parallel delegation where the runtime allows, a verification check at each stage that can actually fail, and a skeptical self-review before delivery.
What it does
The skill shapes the procedure a model follows on complex work. It makes the model decompose before acting, delegate independent sub-work where subagent tooling exists, verify each stage against a failable check rather than a feeling, and critique its own output before delivering it.
What it does not do
It does not change the underlying model's capability. Coherence across long tasks and genuine self-correction live in the model's weights, not in a prompt. On a model that already does these well, the skill reinforces good habits. On a weaker model, it imposes structure the model would otherwise skip, but it cannot raise the reasoning ceiling. Treat it as a checklist, not a capability transplant.
When to use it
Trigger on tasks that span multiple files, multiple sources, or multiple sessions, or when you explicitly ask for systematic execution. Do not use it on tasks with one obvious approach that fit in a single pass. Staging a trivial task wastes effort and buries the answer.
Variants
Three skills share the same core loop. Pick by how you want the work run:
fable-mode- the default. Runs the loop inline on the current model (Opus when that is the host). Use this unless you want the work pinned to a specific model.fable-sonnet- spawns a subagent pinned to Claude Sonnet. The balanced choice: strong reasoning at lower cost than Opus. Requires a runtime with the Agent tool.fable-haiku- spawns a subagent pinned to Claude Haiku. For high-volume or cost-sensitive work where structure matters more than peak synthesis. Requires a runtime with the Agent tool.
The variants pass the same stage map, failable verification, self-critique, warning threshold, and find-and-replace safety rules down to their subagent. They do not raise the chosen model's reasoning ceiling.
Files
SKILL.md- the skill itselfEXAMPLE.md- a worked before/after showing the verification check catching an error that a one-shot attempt shipsfable-sonnet/SKILL.md- the Sonnet variantfable-haiku/SKILL.md- the Haiku variant
Installation
Place each skill directory (fable-mode, and optionally fable-sonnet /
fable-haiku) wherever your Claude environment loads skills from (for example,
a skills directory read by Claude Code), then invoke it by name or let it
trigger on a qualifying task. Each variant's folder name must match the name:
field in its frontmatter.
// compatibility
| Platforms | cli |
|---|---|
| Operating systems | — |
| AI compatibility | claude |
| License | — |
| Pricing | open-source |
| Language | — |
// faq
What is fable-mode?
A Claude skill that activates Fable-style agentic behavior: explicit multi-stage planning, sub-agent delegation, and self-verification.. It is open-source on GitHub.
Is fable-mode free to use?
fable-mode is open-source, so it is free to use.
What category does fable-mode belong to?
fable-mode is listed under skills in the Claudeers registry of Claude-compatible tools.
// embed badge
[](https://claudeers.com/fable-mode)
// retro hit counter
[](https://claudeers.com/fable-mode)
// reviews
// guestbook
// related in Claude Skills
An agentic skills framework & software development methodology that works.
💫 Toolkit to help you get started with Spec-Driven Development
AI coding assistant skill (Claude Code, Codex, OpenCode, Cursor, Gemini CLI, and more). Turn any folder of code, SQL schemas, R scripts, shell scripts, docs,…