Production authorization, built around Cedarling.
Control which tools, data, and actions people and AI agents can access. Use Cedarling for policy decisions, enforce them in your application, and connect identity, policy delivery, and evidence to your existing infrastructure.
Follow a policy store from authoring to evidence
Signed token evidence
Agama Lab Cedar Policy DesignerVersioned .cjar policy store
Cedarling inside the applicationALLOW or DENY → protected operation
Policy stores to Cedarling; logs back to Lock
From policy authoring to authorization evidence
An agent requesting a customer-record export needs an authorization decision before the application returns the data. That decision depends on reviewed policies reaching the runtime through a verified release.
1. Author the policies
Policy author · Agama Lab Cedar Policy Designer
Translate business access rules into a Cedar schema, policies, test requests, and policy-store metadata.
2. Review and publish a release
Reviewers and release workflow · Git and Agama Lab
Review policy changes and test results, then publish a versioned archive with its checksum and signature bundle.
3. Verify the release
Deployment workflow · checksum verification and cosign
Check the archive's bytes and expected publisher identity. Reject the release if verification fails.
4. Distribute and load the store
Lock Server or direct artifact delivery · Cedarling initialization
Deliver the verified archive through Lock or your artifact system, then initialize Cedarling with that store and your bootstrap configuration.
5. Evaluate an authorization request
Application and Cedarling · enforcement and decision points
Send the principal, action, resource, and current context to Cedarling. It returns ALLOW or DENY, a request ID, and available diagnostics.
6. Enforce before the protected operation
Application or gateway · enforcement point
Permit the operation only after ALLOW for that request. A denial or evaluation failure blocks access.
7. Retain the decision record
Application operators · protected logs and Lock Server
Retain decision logs and request IDs for investigation, with sensitive data excluded. Collect signed TRACE evidence where producers support it.
Application enforcement
Application enforcement with Cedarling
Embed Cedarling in your application process or call it through a private sidecar. In both cases, Cedarling is the Policy Decision Point (PDP). Your application or gateway is the Policy Enforcement Point (PEP): it supplies trusted resource data and request context, then permits or blocks the protected operation.
Cedarling integration documentation (opens in a new tab)
CedarlingEnforce the decision before returning protected data or executing an action. Block access on denial, evaluation errors, or an unavailable decision.
The AuthZEN Authorization API (opens in a new tab) defines request and decision exchanges between a PEP and a PDP, such as Cedarling's sidecar.
Trusted identity
Trusted identity with Jans Auth or your existing provider
Keep your existing trusted identity provider, or use Jans Auth for authentication and token issuance. Configure Cedarling's trusted issuers and token mappings so it can validate signed tokens and make their claims available to policies.
Trusted issuer and token validation documentation (opens in a new tab)
CedarlingIf your application establishes identity itself, it can supply application-asserted principals instead. That trusted application boundary is responsible for the identity and facts it sends to Cedarling.
Policy governance
Policy governance with Agama Lab and Git
Use Agama Lab Cedar Policy Designer to author schemas and policies and test authorization requests. Review the source in Git before publishing a versioned policy store. Keep the deployed store's ID, version, and archive hash available for investigation.
Agama Lab policy release documentation (opens in a new tab)
Agama Lab Cedar Policy Designer
CedarlingAgama Lab releases include a .cjar archive, a .cjar.sha256sum checksum file, and a .cjar.sigstore.json signature bundle. Your deployment workflow verifies the release before Cedarling loads it.
Verify bytes and publisher identity
Compare the archive's SHA-256 digest with its checksum file. Then use Sigstore's cosign verify-blob with the signature bundle, the approved signer's identity, and the expected OIDC issuer. Obtain both expected values from trusted release configuration, independently of the downloaded bundle. A checksum alone does not authenticate the publisher.
Verification binds the publisher to the archive's bytes. Your review and tests establish whether those policies meet your business rules. Load the exact verified bytes. Changing those bytes invalidates the signature.
Follow the Agama release instructions (opens in a new tab) and the cosign verification guide (opens in a new tab).
The AuthZEN Policy Store draft (opens in a new tab) proposes a portable format for distributing authorization policies.
Distribution and evidence
Policy distribution and evidence with Lock Server
Use Janssen Lock Server when you need centralized policy-store distribution and decision logs. As a Policy Retrieval Point, it retrieves and caches policy stores for Cedarling instances. You can also load verified archives directly and use your existing protected log pipeline.
Lock Server integration documentation (opens in a new tab)
CedarlingProtect policy retrieval and log delivery with authenticated connections and separate access controls. Set retention rules and exclude sensitive data from logs. Lock also verifies and stores signed TRACE evidence from registered producers.
Signed TRACE evidence
Lock's TRACE support (opens in a new tab) verifies producer signatures, stores accepted assertions unchanged, and lets authorized clients retrieve records by execution or policy-store release. OAuth controls access to the evidence API; record verification uses registered producer keys authorized for particular event kinds.
A PDP signs AUTHORIZATION_DECISION to report its decision and the policy-store ID, version, and archive hash. The enforcement point signs CAPABILITY_INVOKED to report the invocation it mediated. An observer may sign RUNTIME_EFFECT to report an observed effect. Each record requires caller-supplied trace_execution_id and execution_authority to associate it with an execution.
Cedarling server and sidecar builds sign and send AUTHORIZATION_DECISION records to Lock today; the browser (WASM) build does not emit TRACE. Check a signed record in your browser. The TRACE log level controls diagnostic detail; it does not enable signed evidence.
A valid signature identifies the signer and detects changes to signed fields. It does not prove that an assertion is true, that reporting is complete, that enforcement succeeded, or that an action caused an effect.
Integrate Cedarling into your application
Start with the integration interfaces (opens in a new tab) for your runtime, or follow a worked tutorial to see policy decisions and enforcement in a complete application.