AI Decision System
Replayable, offline-verifiable records for AI-assisted go/no-go decisions: experiment promotions, typed model answers and human-approved actions.
Fornax checks what an AI agent tells you against the evidence of what it actually did — VERIFIED, UNVERIFIED, CONTRADICTED, REVIEW, or UNAVAILABLE. No made-up trust score. Currently supports coding ag
Developer toolsTypeScript · JavaScriptGitHub @horonomyOnlineOwner unconfirmed
Have you tried it? One tap is enough. No sign-in needed.
No updates detected yet. GitHub releases will appear here.
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
Fornax marketing/product site (fornax.horonom.com) — problem, how it works, pricing, OSS, security/privacy, docs & app CTAs.
Translation pending
Taken from the repository README
Loading the full README. The saved excerpt is shown below.
The public marketing/product site for Fornax
— evidence-first execution-truth infrastructure for AI agents, currently
supporting coding agents (Claude Code, Codex, opencode). This is not
the technical documentation site (that's fornax-docs); it's the landing
surface that explains the product, presents (placeholder) pricing, calls out
the OSS boundary, and links out to docs, GitHub, and the SaaS app.Read the full README ↗The public marketing/product site for Fornax
— evidence-first execution-truth infrastructure for AI agents, currently
supporting coding agents (Claude Code, Codex, opencode). This is not
the technical documentation site (that's fornax-docs); it's the landing
surface that explains the product, presents (placeholder) pricing, calls out
the OSS boundary, and links out to docs, GitHub, and the SaaS app.
Deploys to fornax.horonom.com (moved from fornax.horo.run per FORNX-328;
fornax.horo.run still aliases to the same deployment but is no longer
canonical — runtime/API surfaces stay on horo.run by design).
package.json/package-lock.json for exact pinned versions)src/index.css (no CSS
framework) — static output, no backendnpm install
npm run dev # start the dev server
npm run build # typecheck (tsc -b) + production build to dist/
npm run lint # oxlint
npm run preview # preview the production build locally
Fornax is pre-1.0 and early. The release shown in the hero badge is
LATEST_RELEASE in src/config.ts: it must equal the latest published
release of horonomy/fornax-core and must never name an unreleased candidate
(npm run check:version verifies this; a daily job also checks the deployed
site). Pricing figures on /pricing are clearly labeled provisional
placeholders, not committed prices; real commercial pricing requires owner
approval. The docs and SaaS app links are configured in src/config.ts
(DOCS_LIVE / APP_LIVE gate them).
Part of the horonomy GitHub organization / Fornax product family. See
horonomy/fornax-core's
docs/adr/ for the architecture and repo/CI conventions this repo follows
(PR-only merge-commit workflow to main, Gitmoji commits, path-filtered CI
with SHA-pinned actions).
MIT — see LICENSE.