EASAPEnterprise agent assurance
Open referenceGitHub ↗Assess an agent
Immediate utility · no account required

Agent Release
Readiness Planner

Describe what your agent can access and do. Get an inherent-risk classification, control gaps, executable scenario pack, named owners, and a machine-readable release gate.

INPUT Agent authority and exposure →OUTPUT Actionable assurance plan
01
Describe the authorityNothing entered here is persisted.
LOCAL INPUT · NO ACCOUNT
Agent context
Deployment stage
Action autonomy
Who is exposed?
Execution volume
Data it can access
Actions it can take
Controls already implemented
Plan current · era_ef8fcddc841f
02
Act on the planDeterministic rules · EASAP-RP-1.0
HIGH RISK
73/ 100
inherent risk
MINIMUM ASSURANCE PROFILE

ENHANCED

CONTROL GAPS OPEN

CONTROL COVERAGE33%
Planning result—not a production approval.The exact integrated agent must still pass its executable assurance campaign.
OPEN CONTROL GAPS60 critical blockers
REQUIRED SCENARIOS624 minimum seeds
REQUIRED APPROVALS4AI engineering · Security · Risk owner · Release authority
WHY THIS CLASSIFICATION
Production deploymentActions require approvalCustomer-facing exposureOperational execution volumeChanges enterprise recordsConfidential data access
IMMEDIATE NEXT DECISIONClose or formally accept 6 control gaps before executing the release gate.
One input · four teams

Leave with work your organization can assign today.

ENGINEERINGCandidate manifest and integration boundary
SECURITYThreat-driven minimum scenario pack
GOVERNANCEControls, owners, approvals, and retention
RELEASE AUTHORITYMachine-readable gate and decision criteria