Start from the work you already trust
AI can read existing Projects, Test Groups, Sections, tags, templates, and cases before it proposes new coverage.
01 / Connect
Let approved AI clients understand your test context, create coverage, prepare Test Runs, execute API checks, and return evidence within the permissions you set.
Move beyond generic answers. Give AI enough context to help with real QA work while Armoury keeps the boundaries visible.
Context, action, and control
Armoury lets an approved client work with real QA records without bypassing the policies, scopes, and Project permissions that govern the team.
AI can read existing Projects, Test Groups, Sections, tags, templates, and cases before it proposes new coverage.
Validate drafts, create and move cases in bulk, prepare Test Runs, add Run Cases, inspect reports, and turn failed results into Defects.
Create API Projects, Collections, Requests, variables, environments, assertions, and modules, then execute them and inspect the history.
Organization policy, OAuth scopes, normal project permissions, write gates, and per-tool audit records control what each client can do.
Authorize before action
Client approval, organization policy, OAuth scopes, and normal permissions are checked before the first tool call and remain in force throughout the task.
Authorize a registered MCP client through KEENLITY ID for the Armoury MCP resource.
Enable MCP for the organization and grant only the read, write, or execute scopes the workflow needs.
The AI uses existing Armoury services and permissions, while every tool action remains available for audit.
Accountable assistance
Your team can move from an idea to structured coverage or an executed API check faster, without moving authorization, tenancy, or auditability outside the system of record.
Review plans and evaluation