fable-mode

fable-mode is a skill for Claude Code, Codex from jtsternberg/claude-plugins. It costs 36 tokens per session (405 once invoked), scanned A, original, MIT.

A working style for testing whether acting like a high-performing frontier AI model improves an agent’s decisions and follow-through. It compares this approach with another operating style under similar conditions.

In plain words
What is it for?
Run A/B evaluations of agent behavior, record results, verify outcomes, and compare judgment, independence, and completion quality.
Why use it?
It provides a structured way to examine whether an agent solves the underlying problem, checks its work, and takes ownership of the result.

Skill for Claude CodeCodex

Part of the jtsternberg plugin — 6 skills, 31 plugins shipped together

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/jtsternberg/claude-plugins/fable-mode
Any agent
npx skills add jtsternberg/claude-plugins --skill fable-mode
Clone the repo
git clone --depth 1 https://github.com/jtsternberg/claude-plugins

Made for: Claude Code, Codex.

Or install jtsternberg, the plugin that ships this one along with the rest of its 6 skills, 31 plugins.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for fable-mode

README.md
[![agentmods](https://agentmods.dev/badge/skills/jtsternberg/claude-plugins/fable-mode.svg)](https://agentmods.dev/skills/jtsternberg/claude-plugins/fable-mode)
Your own site
<a href="https://agentmods.dev/skills/jtsternberg/claude-plugins/fable-mode"><img src="https://agentmods.dev/badge/skills/jtsternberg/claude-plugins/fable-mode.svg" alt="Measured on agentmods" height="20"></a>
Per session 36 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 405 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00036 $0.00405
Opus 5 $0.00018 $0.00202
Sonnet 5 $0.00007 $0.00081
Haiku 4.5 $0.00004 $0.00040

Measured 4d ago against content hash 9febb42e8d9d, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

fable-mode scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 4d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

plugins/codex/skills/fable-mode/SKILL.md · 31 lines

What it actually says

Think Like Fable

Treat Anthropic's Fable as the behavioral benchmark, not as an identity claim. This condition intentionally asks a Terra-class or other Codex agent to operate as though it must match the judgment associated with a top-tier frontier model. Record the task and compare the result with sol-mode under equivalent conditions.

The stance

Own the outcome, not merely the response. Treat the user's message as evidence about the goal rather than a complete boundary for the work. Spend the user's attention carefully: decide reversible matters, and ask only when the choice is genuinely theirs or the action is consequential. Track truth rather than comfort: distinguish what was measured from what was assumed, and verify when challenged.

What this should produce

  • Fix the class of a problem, not only the visible instance.
  • Follow a diagnosis to its consequences, including existing bad state.
  • Consider the shape of the work before starting execution.
  • Verify end to end at the user's actual surface.
  • Make reversible decisions without unnecessary permission loops.
  • Push back when the proposed path is weaker than the evidence-supported path.
  • End with a decision and receipts, not a menu of obvious next steps.

These are manifestations of the stance, not a checklist that substitutes for judgment. When a rule conflicts with the actual goal, reason from the goal and the evidence.

A/B discipline

For a meaningful comparison, hold the task, repository state, available tools, and verification standard constant. Change only the active stance. Capture the initial response, implementation choices, follow-up behavior, verification quality, and user-visible outcome. Do not call a result better merely because it is longer or more forceful.

Do not claim to be Fable or to have access to Anthropic's model. Use the name as the experimental frame.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 4d ago First seen · 31 lines · 36 tokens per session scan A 9febb42e8d9d

Subscribe to this mod's changes

fable-mode is a skill published in the GitHub repository jtsternberg/claude-plugins (5 stars, last pushed today), licensed MIT. It adds 36 tokens to every session and 405 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Related

Other skills, from other repositories

systematic-debugging

Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.

obra/superpowers · 21 tokens

brainstorming

You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation.

obra/superpowers · 37 tokens

auto-perf-optimize

Run agent-driven VS Code performance or memory investigations. Use when asked to launch Code OSS, automate a VS Code scenario, run the Chat memory smoke runner, capture renderer heap snapshots, take workflow screenshots, compare run summaries, or drive a repeatable scenario before heap-snapshot analysis.

microsoft/vscode · 62 tokens

chat-perf

Run chat perf benchmarks and memory leak checks against the local dev build or any published VS Code version. Use when investigating chat rendering regressions, validating perf-sensitive changes to chat UI, or checking for memory leaks in the chat response pipeline.

microsoft/vscode · 51 tokens

chat-pet-sprite-creation

Use when creating or changing VS Code chat pet sprite art, sprite sheets, state animations, eye treatments, Stable/Insiders variants, or pet transitions under src/vs/workbench/contrib/chat/browser/widget/media/chatPet.

microsoft/vscode · 53 tokens

cpu-profile-analysis

Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…

microsoft/vscode · 71 tokens