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. In a public benchmark, about 38% cheaper than Claude Sonnet 5 on equal footing and as good or better in 74% of 50 tasks it had never seen. It gives each task to the cheapest model that do
Developer toolsPython · HTMLGitHub @cintocasalsOnlineOwner 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
Frontier prices only for frontier work: each task goes to the cheapest model that does it well, a request is split only when that pays, and every result is checked. Decisions by Jev.
Taken from the repository README
Loading the full README. The saved excerpt is shown below.
Frontier prices only for frontier work.
Taskpenny splits a request only when it expects splitting to pay, gives each task to the cheapest model that does
it well, checks every result, and shows you the receipt. Open source, Python, one command to try.
A real run replayed at 4x. Watch it in your browser: live demo
(no key, no cost). The saving on the page is the receipt's estimate against Claude Opus 5.5; the benchmark below
runs the models it compares with for real.Read the full README ↗The full README is unavailable right now. Read it at the original link. The saved excerpt is shown below.
Frontier prices only for frontier work.
Taskpenny splits a request only when it expects splitting to pay, gives each task to the cheapest model that does
it well, checks every result, and shows you the receipt. Open source, Python, one command to try.
A real run replayed at 4x. Watch it in your browser: live demo
(no key, no cost). The saving on the page is the receipt's estimate against Claude Opus 5.5; the benchmark below
runs the models it compares with for real.Read the full README ↗