RuleFence documentation
Understand the boundary.
Build against it deliberately.
Trace every request from managed identity to WordPress authority, policy, approval, execution, evidence, and recovery.
CORE MODEL
Every request follows the same explicit path.
Connection begins the request. Governance decides what happens next—and leaves the result inspectable.
Quick start
Reach the first inspectable decision.
Use one managed agent and one narrow Ability to learn the complete path before adding consequence.
Install
Activate the local control plane and confirm the site is ready.
Register
Give the agent a managed identity tied to its purpose and environment.
Grant
Allow one explicit WordPress Ability with the smallest useful scope.
Inspect
Read the decision, outcome, evidence, and recovery path.
Documentation map
Learn the model in the order requests use it.
The documentation follows the operating path rather than separating related controls into disconnected features.
Foundations
Know the boundary
Decisions
Control each request
Operations
Keep outcomes inspectable
Developer reference
Declare what the agent may request. Never invent authority.
An integration should make identity, the requested Ability, context, and expected evidence explicit. WordPress authorization and policy remain separate gates.
- Name the managed identity
- Request a registered WordPress Ability
- Provide the context policy needs
- Preserve the resulting evidence
Troubleshoot the path
Begin where the request stopped.
Do not debug the agent, WordPress, and policy as one opaque system. Identify the gate, then inspect its evidence.
Agent not recognized
Confirm the managed identity, environment, and registration state.
Identity troubleshootingWordPress denied access
Inspect the user, role, capability, and native authorization result.
Authorization troubleshootingRequest blocked or paused
Read the risk context, matching rule, and approval requirement.
Policy troubleshootingExecution needs recovery
Trace the result, linked evidence, and available recovery control.
Recovery troubleshootingReference and help
Keep documentation beside the operating record.
Use the knowledge base for focused answers, the changelog for behaviour changes, and support when the complete request context needs review.
Begin with the complete path
Understand one request.
Then expand deliberately.
Identity, authority, policy, approval, evidence, and recovery stay connected.