Claude Fable 5 Suspended: Skills That Still Work on Opus 4.8

On June 12, 2026, the US government issued an export control directive and Anthropic pulled Claude Fable 5 and Mythos 5 offline for every user worldwide. As of writing, both models remain suspended with no confirmed restoration date.

The good news: every other Claude model is unaffected. Opus 4.8, Sonnet 4.6, Haiku 4.5, Claude Code, the API, everything works. And all your SKILL.md skills work exactly the same on Opus 4.8 as they did on Fable 5.

Quick Answer: Claude Fable 5 is suspended under US export controls since June 12, 2026. Switch to Opus 4.8. All SKILL.md skills work on it. Skills actually matter more now because they help close the capability gap between Opus 4.8 and Fable 5.

What happened

Fable 5 launched June 9. Three days later, the US Commerce Department issued a directive ordering Anthropic to suspend access for all foreign nationals worldwide. Since Anthropic can't verify nationality on every API request, they disabled both models for everyone. The stated trigger was a jailbreak concern. Anthropic publicly disagrees and says it's a misunderstanding.

Recommended skills

skill-router-2

by Shippers · 3

Automatically detect, load, and stack the perfect skills combo for any user requ…

$5

Solo SaaS Architect

by tudor.ai

Automatically builds complete, launch-ready SaaS websites, databases, and secure…

$5

Auto-Approve Discipline Kit

by Zotri.11

Run Claude Code unattended with a battle-tested safety framework, hardened deny-…

$19

Browse all Claude Fable 5 skills

Why skills matter more now

Fable 5 was better than Opus 4.8 on hard coding tasks. The SWE-bench Pro gap is real: 80.3% vs 69.2%. For long-horizon autonomous work, large migrations, and multi-file refactors, Fable 5 had an edge.

Skills help close that gap. A well-written skill gives Opus 4.8 the domain knowledge and procedural discipline it needs to perform closer to Fable 5 on specific tasks. Microsoft Research actually proved this with SkillOpt: optimized SKILL.md files lifted model performance by 19-24 points across benchmarks. The skill compensates for what the model lacks.

Skills to install right now

Code Reviewer. Opus 4.8 is slightly more likely to miss edge cases than Fable 5 was. A code review skill catches what the model misses. This is the most important skill to have active.

Token efficiency skills. Opus 4.8 uses fewer tokens than Fable 5 per task, but if you're on a Max plan you want to maximize what you get. Skills that reduce unnecessary tool calls and keep the agent focused save tokens.

Testing skills. Fable 5 was better at generating comprehensive test suites autonomously. With Opus 4.8, a testing skill that specifies your patterns, frameworks, and coverage expectations closes the gap.

Architecture and planning skills. Fable 5 could hold larger plans in context. Opus 4.8 benefits from skills that break complex tasks into phases and enforce structured thinking.

Security audit skills. Don't skip security review just because you're on a less capable model. If anything, run security skills more aggressively on Opus 4.8 output.

Browse all Claude Code skills at agensi.io/skills.

How to switch

If your code references claude-fable-5, change it to claude-opus-4-8. That's the only change needed. Claude Code automatically defaults to Opus 4.8 for new sessions. Your skills, hooks, CLAUDE.md, and project configuration all work the same.

When will Fable 5 come back?

Unknown. As of late June 2026, Anthropic is negotiating with the administration. A bipartisan House letter demanded the directive's legal justification. Trump signaled positivity at the G7 summit. Polymarket had 58-67% odds of US restoration by July 1. But nothing is confirmed.

Plan as if Opus 4.8 is your model for the next month. Build your skill stack around it. If Fable 5 returns, your skills will work even better on it.

Disclosure: I run Agensi, a marketplace for security-scanned SKILL.md skills.

---

Related: Best Claude Code Skills 2026 and How to Install Skills in Claude Code.