Run and observe
The rabbinatix CLI
Every rabbinatix command: what it does, what it sends, the token it needs and how it ends, for your terminal and your CI.
Open API & CI in the appInstall and connect
rabbinatix is one npm package with one command. It needs Node.js 20 or later, on Linux, macOS or Windows. Run it with npx, so CI always gets the latest:
npx -y rabbinatix --helpIt finds Rabbinatix in this order: the --api option, then the RABBINATIX_API environment variable, then "api" in .rabbinatix/config.json in your repo. The token comes from RABBINATIX_TOKEN (or the variable --token-env names). Never put a token in the config file: the CLI refuses a config file that holds one. Make tokens under Integrations → API tokens; API & CI shows the address to use.
Every command follows the same rules:
- Fixed rules: the same input gives the same answer, so it can block a merge or a deployment. It never calls an AI model.
- Your code stays on your machine. Each command's
--helpends with what it sends. - Exit codes:
0done,1the check you asked for failed (--fail-on),2an error. --format jsonprints the whole answer for scripts;--out <file>also writes it.- Where a result needs understanding, it ends with the AI prompt to give your assistant. Which does what: AI assistant or CLI.
- You don't have to remember the options: each page has a button that builds the command for its job, with your API address, and the token step it needs. Scan the code on Security checks, Keep the code in step on Logic, Run the test cases on Flows, Put code on the map on Systems, Check before deploying on Deployments, Run it on my computer on Test server, Send from my computer on Test APIs, Build it in the screen editor, Check real calls in a system's *…* menu, and Check a change on Code impact. Each also builds the prompt for your AI assistant, where it has one.
Commands
rabbinatix impact
What a code change affects on the map: the logic its code belongs to, the flows and contracts around it, and the test cases to run. A pull request, your work in progress, or any two branches or commits (--base main --head feature/x, v1.2.0..v1.3.0).
npx -y rabbinatix impact --fail-on breaks --comment impact.md --out impact.json- Sends: Which anchors, routes and spec operations the change touches, and the changed file names. No code.
- Token:
read(deployto keep the report) - Exit code: 1 when the change may break a contract (
--fail-on) - AI prompt it prints: Fix what this change broke · Put the changed code on the map
rabbinatix status
Whether your code is in step with the map: anchors behind it, anchors whose logic is gone, logic with no anchor, and files with API routes but no anchor. --write-map rewrites .rabbinatix/code-map.json from the anchors.
npx -y rabbinatix status --fail-on behind,gone- Sends: The anchors' keys and revisions, with their file names and lines. No code.
- Token:
read - Exit code: 1 when
--fail-onfinds something - AI prompt it prints: Bring my code in step with every change · Put my repo on the map
rabbinatix test
Runs the flows' test cases, simulated, as Run does in the app: test data answers every call and nothing reaches the network. --affected impact.json runs only those a change affects.
npx -y rabbinatix test --affected impact.json- Sends: The flow keys and system IDs you name.
- Token:
read - Exit code: 1 when a test case fails (
--fail-on) - AI prompt it prints: Find out why these test cases fail
rabbinatix scan
Security issues in your repo: secrets, risky code and vulnerable dependencies, with Semgrep, Gitleaks and OSV-Scanner too when they are installed. --report sends the findings to Code scan.
npx -y rabbinatix scan --fail-on high --report- Sends: Nothing, unless you add
--report: then each finding's rule, file, line and a short description. No code. - Token:
scan(only with--report) - Exit code: 1 when a finding is at least as severe as
--fail-on
rabbinatix deploy check
Whether a release may go to an environment, from its contracts: the same check as the Safety check view on Deployments.
npx -y rabbinatix deploy check --system orders --version 1.4.0 --env prod- Sends: The system, the version and the environment.
- Token:
read - Exit code: 1 when the deployment would break a contract
- AI prompt it prints: Explain why this deployment is blocked
rabbinatix deploy record
Records a deployment after your deploy step. An environment that asks for approval gets a promotion request instead.
npx -y rabbinatix deploy record --system orders --version 1.4.0 --env prod- Sends: The system, the version and the environment.
- Token:
deploy - Exit code: 0 when recorded or requested
rabbinatix generate
The code of some logic or a whole system, in TypeScript, Python, C#, Java, Go, Kotlin, Swift, Dart, PHP, Ruby or Rust, as the Code panel downloads it. --check writes nothing and fails when the repo's generated files are missing or older than the logic. --record lists the generation on the code panel's Exported tab, as a download is.
npx -y rabbinatix generate --system orders --language python --check- Sends: The logic keys or the system, and the language. With
--record, each written file's name and fingerprint. - Token:
read - Exit code: 1 when
--checkfinds files to regenerate
rabbinatix verify
Builds the Python, C#, Java, Go, Kotlin, Swift, Dart, PHP, Ruby or Rust code of a logic on your machine and replays its latest recorded test runs through it: the same answers, calls and path, or the first difference. Needs Python 3.10+, .NET 8 SDK, JDK 21, Go 1.22+, Kotlin 1.9+ (kotlinc) with a JDK, Swift 5.9+, Dart 3+, PHP 8.1+ with mbstring, Ruby 3.1+ or Rust 1.70+ (rustc).
npx -y rabbinatix verify --logic LOGIC-ORDERS-PLACE-001 --language java- Sends: The logic key and the language.
- Token:
read - Exit code: 1 when a test run differs
rabbinatix openapi pull
A system's OpenAPI file, stamped with its version, for client generators in any language.
npx -y rabbinatix openapi pull --system orders --out orders.yaml- Sends: The system and the version.
- Token:
read - Exit code: 0
rabbinatix openapi diff
What your repo's OpenAPI file changes against the system's APIs on the map: operations added or removed, and fields changed, with which would break callers.
npx -y rabbinatix openapi diff api/openapi.yaml --system orders- Sends: The file's operations and schemas, not its descriptions, servers or other text.
- Token:
read - Exit code: 1 when a change would break callers (
--fail-on)
rabbinatix design
A screen's design in a folder, to build it pixel-perfect by hand or with your AI assistant: the page as plain HTML with its styles, the theme as design tokens, each widget's key and type, and, when a picture of the screen is stored, the picture and every widget's box and styles. A picture is stored when the screen goes on a whiteboard, when someone downloads its design pack, and at each release of its system.
npx -y rabbinatix design SCREEN-WEB-LOGIN-001 --width phone- Sends: The screen's key, the width and whether its layout goes around it.
- Token:
read - Exit code: 0
rabbinatix mock
A system's test server on your machine, answering from its examples.
npx -y rabbinatix mock --system auth-service- Sends: The system you name.
- Token:
read - Exit code: 0 when stopped
rabbinatix relay
Sends Test APIs' requests from your computer, so the API needs no CORS headers and local or VPN addresses work. Listens on this computer only and needs the code it shows.
npx -y rabbinatix relay- Sends: Nothing to Rabbinatix: each request goes straight to the API it names.
- Token: none
- Exit code: 0 when stopped
rabbinatix watch import
Checks the calls in a HAR recording or a file of OpenTelemetry traces against a system's contracts, as the contract watcher does in the browser, and reports them the same way.
npx -y rabbinatix watch import session.har --watcher src/dev/rabbinatix-watch.js- Sends: Per finding: the API, what was expected and what was found, and how often. Per call: the API, status and timing. Never the values seen.
- Token: the watcher's own key (in the downloaded file)
- Exit code: 1 when a finding is at least as bad as
--fail-on
rabbinatix watch listen
Checks the calls a running app sends it against a system's contracts: phone apps through the files in the watcher download (Android, iOS, Flutter, React Native), and services through OpenTelemetry. Prints each new finding and reports them as the watcher does. Listens on this computer only unless you add --host 0.0.0.0, which then needs the code it shows.
npx -y rabbinatix watch listen --watcher rabbinatix-watch.js- Sends: Per finding: the API, what was expected and what was found, and how often. Per call: the API, status and timing. Never the values seen.
- Token: the watcher's own key (in the downloaded file)
- Exit code: 0 when stopped (1 when a finding is at least as bad as
--fail-on)
In a pull request and a deploy
A workflow that runs on every pull request, then gates and records each deploy. Integrations → API & CI has it ready to copy with your address filled in.
On every pull request:
npx -y rabbinatix scan --fail-on high --report
npx -y rabbinatix impact --fail-on breaks --comment impact.md --out impact.json
npx -y rabbinatix status --fail-on behind,gone
npx -y rabbinatix test --affected impact.jsonOn main, for each environment:
npx -y rabbinatix deploy check --system orders --version "$VERSION" --env prod
./deploy.sh
npx -y rabbinatix deploy record --system orders --version "$VERSION" --env prodWhere the environment requires sign-off from affected teams, deploy record is refused until they sign off. --allow-missing-signoff "hotfix for the outage" moves it anyway and records the reason.
Related
- AI assistant or CLI: which does whatWhere your AI assistant works and where the rabbinatix CLI works, the rules between them, and every feature that touches code.
- CI and the APIGate deployments in CI with the safety check and record what you deploy, using an API token and the HTTP API.
- Code impact: what a change affectsCheck a code change against the map: a pull request in your pipeline, any two branches or commits, or your AI assistant's own work. Which logic, flows and contracts it reaches, which test cases to run, and whether the map still says what the code does.
- Security checksFind the usual API security mistakes in your design and your code: an API that changes data without sign-in, a password in a response, a token in the URL.
- Programming languagesWhat works in which programming language: anchors and highlighting in all of them, plus code generation, code impact, API routes, security rules and code snippets.
