AI Decision System
Replayable, offline-verifiable records for AI-assisted go/no-go decisions: experiment promotions, typed model answers and human-approved actions.
Open-source workflows that make context, scope, validation, evidence, and human review explicit for AI coding agents — on your next pull request, and across coordinated fleets. Plus practical adoption
Developer toolsRuby · ShellGitHub @shakacodeOnlineOwner unconfirmed
Have you tried it? One tap is enough. No sign-in needed.
Owners of public repositories under a personal GitHub account can sign in to verify ownership and manage the introduction themselves. No administrator approval is needed.
Sign in, generate a verification file, and commit it to this repository’s default branch within 30 minutes. Files on other branches or forks do not count. Then check the file below to manage the project without administrator approval. Keep the file to retain your access.
If needed, you will go to sign-in. After signing in, return here and press your chosen verification button again.
Run /nomorevibe verify in your project folder to verify ownership and manage the introduction yourself.
Verify in one minuteAsk about other verification options
Give your agent a task. Get a verified, explained pull request.
Translation pending
Taken from the repository README
Loading the full README. The saved excerpt is shown below.
Give your coding agent a task. Get a tested, reviewed PR that's easy to understand.
Shaka guides the work from the first question through implementation, local tests,
independent review, and delivery on GitHub. You spend less time directing the
process and checking whether the agent finished the job.
Shaka checks for existing work, considers whether the change is worth doing, and
recommends a model and effort level. You can also give it an issue or task link:Read the full README ↗Review the intent and the evidence, not every line.
Give your coding agent a task. Shaka guides it through implementation, verification, independent review, and PR delivery. Each PR shows:
Then you approve the change, request revisions, or resolve an open question.
The agent builds, verifies, and reviews, fixing what it can. Then you review one clear PR: what changed, how it was checked, and what is still open. At the checkpoint, decide what happens next.
Install Shaka, then give it a task:
$shaka <task or issue link>
Examples use Codex. Use /shaka in Claude Code, Cursor, or OpenCode; in Pi,
load the installed skill. You can describe an outcome or provide an issue link.
If setup is missing, Shaka asks whether you want individual or team settings,
finds your existing checks, and defaults to Ask.
Go starts with your agent's current model and effort settings.
Working with Shaka covers practical prompts,
merge choices, feedback, and resuming work.
New to branches and pull requests? Shaka in pictures explains them and walks you through a practice task.
See what a Shaka PR gives you.
Ruby 3.4 or later, Git, an authenticated GitHub CLI, and a coding agent that can load skills and run commands. The installation prompt lets your agent check these for you.
See what to expect from Shaka for setup choices, verification gaps, and help resuming work when checks or access fail.
The MIT license covers the code and documentation. It does not grant rights to use the Shaka or ShakaCode names or logos. ShakaCode is a registered trademark of ShakaCode LLC in the United States. You may say that your work uses or is based on Shaka. Give a fork or product its own name, so it does not appear to come from ShakaCode.