Check which of your repositories Fable 5 will actually cooperate on before spending your usage quota.
Get a cost estimate in dollars and percent of your Claude plan before running an analysis across many repos.
Receive a free architecture review of each scanned repository as a side effect of the check.
Decide per component whether to route work to Fable or to the more capable Opus model.
| botable-dev/scoperoute | 0marildo/imago | 0xdfi/glm-5.2-1m-4x-dgx-spark | |
|---|---|---|---|
| Stars | 3 | 3 | 3 |
| Language | Python | Python | Python |
| Setup difficulty | easy | easy | hard |
| Complexity | — | 2/5 | 5/5 |
| Audience | developer | general | researcher |
Figures from each repo's GitHub metadata at analysis time.
Fable probing requires an account with at least 30 day data retention or requests fail.
Scoperoute is a command line tool and Claude Code skill that helps you figure out which of your code repositories can actually be built using the Claude Fable 5 model, and which ones should be routed to the more expensive Opus model instead. The problem it solves is that Fable 5 sometimes quietly refuses certain repos based on their real code, silently falling back to Opus, so you might think you are using the cheaper or free model when you are not. When you run it against your projects, scoperoute reads through each repository's files, summarizes its components, and then actually tries asking Fable to work on the real code, not just a description of it. Based on how Fable responds to each part, it gives you a verdict per project, such as fully Fable friendly, partially blocked on a sensitive component, or genuinely sensitive across the board. It also estimates the cost of running this check in tokens, dollars, and as a percentage of your Claude subscription plan, so you know what you are spending before you commit to it. As a side benefit, every scan leaves behind a short architecture review of each scanned repository, written by Fable, so the money and time spent on the check also produces something useful for your project. It installs with a single script that sets up both a terminal command and a Claude Code skill, and does not require any separate API key beyond your existing Claude access. It only reads metadata and verdicts about your code for its reports, and does not include your actual source code in the output. This tool is meant for developers who already use Claude Code across multiple repositories and want to know in advance which model to use on each one before spending their usage quota.
A tool that tests your repos against Claude Fable 5 to tell you which ones it will actually build versus quietly fall back to the pricier Opus model.
Mainly Python. The stack also includes Python, Claude Code, Claude Fable 5.
Setup difficulty is rated easy, with roughly 5min to a first successful run.
Mainly developer.
This repo across BitVibe Labs
Verify against the repo before relying on details.