Authorization Policy

Cedar-based authorization policies for Spice.ai Enterprise — fine-grained access control over datasets, models, tools, and endpoints.

Spice.ai Enterprise enforces fine-grained authorization using Cedar, the open-source policy language developed by AWS. Policies decide, on every request, whether an authenticated principal may perform a given action on a Spice resource.

Cedar-based authorization policy is a Spice.ai Enterprise feature. It is layered on top of Authentication — authentication establishes who the principal is; policy decides what they can do.

Authorization Model

Policy evaluation is the standard Cedar (principal, action, resource, context) decision. Spice.ai supplies an embedded Cedar schema that defines the entity types, actions, and attributes available in policies.

Entity types

Entity	Description	Notable attributes
Spice::User	An authenticated principal (OIDC subject, API key identity, …). in [Role]	org_id
Spice::Role	A role the user belongs to. For OIDC principals, sourced from the configured role/group claims; for API-key principals, the key's read or read_write permission level	—
Spice::Dataset	A registered dataset (table) in the runtime	catalog, schema
Spice::Model	An LLM model available for inference	—
Spice::Tool	A tool (built-in or MCP) available for execution	—
Spice::Endpoint	An API endpoint category (e.g. chat, search, sql)	—

Actions

Action	Applies to	Description
Spice::Action::"query"	Spice::Dataset	SELECT / scan of a dataset.
Spice::Action::"read"	Spice::Dataset	Reading a dataset's contents. Carries the fine-grained row filter and column mask annotations (see Row Filters and Column Masks). A read permit also implicitly authorizes query.
Spice::Action::"insert"	Spice::Dataset	INSERT write path.
Spice::Action::"update"	Spice::Dataset	UPDATE write path.
Spice::Action::"delete"	Spice::Dataset	DELETE write path.
Spice::Action::"ddl"	Spice::Dataset	DDL operations on a dataset.
Spice::Action::"invoke"	Spice::Model	Inference / chat completion.
Spice::Action::"execute"	Spice::Tool	Tool invocation (built-in or MCP).
Spice::Action::"access"	Spice::Endpoint	Reaching an endpoint category.

The Cedar schema is fixed by the runtime; policies reference these types and actions directly.

Configuration

Policies are configured under runtime.authorization in spicepod.yaml. Authentication must be configured separately under runtime.auth — see Authentication.

runtime:
  auth:
    oidc:
      issuer_url: https://auth.example.com/
      audience:
        - spice-runtime

  authorization:
    enabled: true              # default: true
    default: allow             # allow | deny  — decision when no policy matches
    provider: local            # local | operator | cloud
    policies:
      - name: analysts-read-only
        cedar: |
          permit(
            principal in Spice::Role::"analyst",
            action == Spice::Action::"query",
            resource
          );
      - name: block-pii
        path: ./policies/block-pii.cedar

runtime.authorization fields

Field	Type	Default	Description
enabled	bool	true	Toggles policy evaluation.
default	allow | deny	allow	Decision when no policy matches. Set to deny for production deny-by-default deployments.
provider	local | operator | cloud	local	Source of policies. local reads inline / file; operator polls the Spice K8s Operator; cloud polls the Spice Cloud Management API.
policies	list of PolicyDefinition	[]	Policy entries (used with provider: local).
operator	object	—	Operator provider config. Fields: endpoint, poll_interval (default 30s).
cloud	object	—	Cloud provider config. Fields: poll_interval (default 60s).

PolicyDefinition

Field	Type	Description
name	string	Human-readable identifier used in logs and decision diagnostics.
cedar	string	Inline Cedar policy text. Mutually compatible with path; both may be used together.
path	string	Path to a .cedar file, resolved relative to spicepod.yaml. Mutually compatible with cedar.

Policy Examples

Default-deny baseline

runtime:
  authorization:
    default: deny
    policies:
      - name: admins-everything
        cedar: |
          permit(
            principal in Spice::Role::"admin",
            action,
            resource
          );

Read-only analysts

permit(
  principal in Spice::Role::"analyst",
  action == Spice::Action::"query",
  resource
);

forbid(
  principal in Spice::Role::"analyst",
  action in [
    Spice::Action::"insert",
    Spice::Action::"update",
    Spice::Action::"delete",
    Spice::Action::"ddl"
  ],
  resource
);

Restrict a dataset to a single role

forbid(
  principal,
  action,
  resource == Spice::Dataset::"sales.public.orders"
)
unless { principal in Spice::Role::"sales-ops" };

Block a PII schema from non-privileged roles

To mask individual columns rather than block access entirely, use column masks.

forbid(
  principal,
  action == Spice::Action::"query",
  resource
)
when {
  resource.catalog == "customers" &&
  resource.schema == "pii" &&
  !(principal in Spice::Role::"compliance")
};

Limit model invocation to a paid tier

permit(
  principal in Spice::Role::"premium",
  action == Spice::Action::"invoke",
  resource == Spice::Model::"gpt-4o"
);

Endpoint-level access (e.g. lock down /v1/sql)

forbid(
  principal,
  action == Spice::Action::"access",
  resource == Spice::Endpoint::"sql"
)
unless { principal in Spice::Role::"engineer" };

Policy Providers

Provider	Use case	Reload
local	Inline Cedar text and/or .cedar files in the spicepod.	Reloaded on Spicepod restart or operator-driven config rollout.
operator	Centralized policy distribution via the Spice K8s Operator.	Polled at poll_interval (default 30s); engine atomically swaps the set.
cloud	Centralized policy distribution via Spice Cloud.	Polled at poll_interval (default 60s); engine atomically swaps the set.

Policy reloads are atomic: in-flight requests complete against the previous policy set, and subsequent requests evaluate against the new set. Empty policy fetches do not silently disable enforcement — combined with default: deny, an empty set denies everything.

Row Filters and Column Masks

Beyond coarse allow/deny, policies enforce fine-grained access — row-level filtering and column-level masking — on the read action. The runtime compiles these from Cedar policy annotations into SQL that is applied to the table scan, so filtered rows and masked values are never materialized for the request: they are enforced before any downstream operator (or the client) observes them.

Annotations attach to a permit policy for Spice::Action::"read". A read permit also implicitly authorizes query, so a single read policy can grant masked, row-filtered access in one rule.

Annotations

Annotation	Effect
@row_filter("<sql predicate>")	Adds a SQL boolean predicate as a row filter. May be repeated with a suffix (e.g. @row_filter_region); multiple filters are AND-combined.
@mask_<column>("<sql expr>")	Replaces <column> with the SQL scalar expression.
@column_mask_<column>("<sql expr>")	Equivalent to @mask_<column>.
@column_mask("<column>=<sql expr>")	Equivalent, with the column named in the annotation value.
@mask_tag_<tag>("<sql expr>")	Replaces every column carrying <tag> (see Tagging columns).
@column_mask_tag("<tag>=<sql expr>")	Equivalent, with the tag named in the annotation value.
@target_table("<dataset>")	Scopes the annotations to a single dataset when the policy matches more than one.

Row filters and mask expressions are evaluated with the request's identity, so they can reference the identity SQL functions (current_user_id(), current_org_id(), current_user_has_role('...')) for per-user or per-tenant access.

Example

// Physicians: full access to patients in their own organization.
@id("physician_read")
@row_filter("org = current_org_id()")
permit(
  principal in Spice::Role::"physician",
  action == Spice::Action::"read",
  resource == Spice::Dataset::"patients"
);

// Analysts: same org scoping, but SSN and any PHI-tagged column are masked.
@id("analyst_read")
@row_filter("org = current_org_id()")
@mask_ssn("'XXX-XX-XXXX'")
@mask_tag_phi("'REDACTED'")
permit(
  principal in Spice::Role::"analyst",
  action == Spice::Action::"read",
  resource == Spice::Dataset::"patients"
);

With these policies, an analyst running SELECT * FROM patients sees only their organization's rows with ssn and any PHI-tagged column replaced, while a physician sees their organization's rows unmasked — from the same query.

Tagging columns

Tag-based masks (@mask_tag_* / @column_mask_tag) target columns by the dataset's column metadata:

datasets:
  - from: postgres:patients
    name: patients
    columns:
      - name: diagnosis
        metadata:
          tags:
            - phi

Rules

• Fine-grained annotations are honored only on permit policies. Attaching one to a forbid is a load-time error.

• A row filter must evaluate to a Boolean; a column mask expression must return the column's data type. Type mismatches fail closed — the query errors rather than returning unmasked data.

• If two matching policies define conflicting masks for the same column or tag, policy load fails.

• To deny access outright, use a forbid on read/query rather than a mask.

The identity SQL functions are also available directly in dataset views and ad-hoc WHERE clauses for row-level logic outside of policy enforcement.

Distributed Cluster Behavior

In a SpicepodCluster, the scheduler is the source of truth for policy. Executors pull policy as part of the GetAppDefinition bootstrap RPC and re-evaluate when policy changes are pushed; the same Cedar decision applies regardless of which scheduler routes a request. Policy diagnostics (matched policy IDs, decision) are emitted as structured logs and surface in runtime.task_history.

Production Checklist

• runtime.authorization.enabled: true and default: deny for production deployments.

• Roles are sourced from a trusted identity provider (OIDC groups or API key tags) — see Authentication.

• Policy files are stored in Git and rolled out via GitOps (Argo CD / Flux) rather than ad-hoc edits on the cluster.

• An "admin escape hatch" permit rule exists and is restricted to a small Spice::Role::"admin" group with audited membership.

• Policy decision logs are forwarded to the SIEM together with authentication logs.

• Cedar policies are validated in CI with cedar validate against the embedded Spice schema before merge.

See also

• Authentication

• Security

• Distributed Query

• Cedar policy language reference