---
title: "Workday Deployment Agent vs. Kiweely Execution Agent"
description: "As Workday documents it, Deployment Agent inspects non-Production tenants and returns setup steps; Kiweely Vine reads and executes Workday changes with consent."
url: https://kiweely.com/blog/workday-deployment-agent-vs-execution-agent
---

# Tenant Inspection vs. Tenant Execution: Where Workday’s Deployment Agent Stops and Kiweely Starts

Documented as read-only, Workday’s Deployment Agent inspects Implementation, Sandbox, and Preview tenants and suggests manual steps; change-making is in partner early access. Kiweely Vine executes approved tenant changes, builds Extend apps and Studio integrations, and browser-tests workflows.

-   By Kiweely
-   November 13
-   Updated November 19
-   7 min read

As Workday documents it, Workday’s Deployment Agent stops at read-only diagnosis: even with the **Ask** skill and **Tenant Inspection** enabled, it only reads Implementation, Sandbox, and Preview configurations and tells you which Workday task to open by hand. (With [Configuration Cloud](https://blog.workday.com/en-us/configuration-cloud.html), in opt-in early access for Workday partners since 18 Nov 2026, Deployment Agent can also carry out changes once approved.) [Kiweely Vine](https://kiweely.com/vine) handles both inspection and execution on your desktop, reading live tenant metadata and—with your explicit consent—configuring calculated fields, reports, business processes, security policies, EIB loads, Workday Extend apps, and Workday Studio integrations.

For implementation consultants and HRIS administrators, that boundary determines whether an AI tool shaves five minutes off documentation lookup or compresses a two-day build-and-test cycle into an afternoon.

## Two Different Workday Deployment Agents: Customer Central Q&A vs. ASOR Ask Skill

Workday uses the Deployment Agent name for two distinct experiences with different contract requirements, security setups, and costs.

The baseline version is **Deployment Agent (Q&A)** inside Customer Central, controlled through **Maintain Feature Opt Ins**. It costs nothing and requires neither a Universal Main Subscription Agreement (UMSA) nor Agent System of Record (ASOR) setup, because it never inspects your tenant. It answers conceptual questions using Administrator Guide topics and Workday Community articles, and is currently limited to US and EU data centers (excluding Canada, Australia, the UK, Japan, and Federal environments).

The tenant-aware version is the ASOR-registered **Deployment Agent**, available in Implementation, Sandbox, Preview, and connected Customer Central tenants once your organization is on UMSA. Enabling it in each non-Production tenant requires a multi-step security bootstrap:

1.  Run **Create Security Policy for Domain** on **Deployment Agent** (if not yet created), then open **View Domain** for **Deployment Agent Users** and grant **View** report/task permissions to groups such as **Implementers**.
2.  Run **Activate Pending Security Policy Changes**.
3.  Open **Agent Management Hub**, locate **Deployment Agent** on the **Unregistered Workday Agent** tab, and click **Register**.
4.  Click **Configure Agent**, toggle on the **Ask** skill (which bundles two underlying ASOR skills: **Tenant Inspection** and **Implementation Knowledge**), populate the **Available To** interaction policy with security groups that also belong to **Deployment Agent Users**, and click **Activate**.

Non-Production Flex Credit Billing Trap

Unlike most Workday-built agents that only meter Production activity, the ASOR **Deployment Agent** bills usage in Implementation, Preview, and Sandbox environments directly against your Production Flex Credit balance. You cannot run the agent in Production at all; configuring it in Production only preserves the setup across Sandbox refreshes.

## What Read-Only Tenant Inspection Actually Returns

When you prompt the ASOR Deployment Agent with the **Ask** skill active, Workday grounds its answer by searching your tenant, opening read views, and checking configuration metadata. That context is genuinely useful when you inherit an unfamiliar tenant and want to know which domain secures a custom compensation field or how your management hierarchy is ordered.

Where the workflow stalls is the handoff back to manual data entry. Consider a routine implementation request: categorizing workers into `Manager` vs. `Individual Contributor` based on your tenant’s management levels so you can drive condition rules and summary reports.

text

```
What Deployment Agent (Ask skill) does:
1. Reads your tenant's Management Level hierarchy (e.g., Lead, Director, VP).
2. Prints text instructions:
   - Open Create Calculated Field
   - Business Object: Worker
   - Function: Evaluate Expression (Default: Individual Contributor)
   - Condition: Management Level In Selection List (Lead, Director, VP...) -> Manager
3. Stops. You build the Sub-Calculation condition, build the Evaluate Expression field,
   attach it to your report or BP condition rule, and test it manually.
```

The same pattern holds for security troubleshooting. If a Payroll Specialist cannot see a custom field, Deployment Agent can inspect the tenant, identify **Worker Data: Compensation Explicit** as the governing domain, and note that **Payroll Specialist** lacks **View** permission. You still open **Edit Domain Security Policies**, update the permission grid, run **Activate Pending Security Policy Changes**, proxy in as a worker in that role, and verify that constrained organization boundaries behave as expected.

If you are building a single calculated field, typing the parameters yourself takes three minutes. If you are replicating 34 eligibility condition rules and their nested Extract Single Instance, Lookup Related Value, and Evaluate Expression chains across tenants—or building an inbound EIB, a custom Extend app, or a Studio integration—the read-only Ask skill leaves 90% of the keyboard work on your plate.

## Capability Comparison: Deployment Agent vs. Kiweely Vine

The architectural difference comes down to where each agent runs and what Workday interfaces it can drive. Deployment Agent lives inside non-Production Workday UI tabs with read-only ASOR tools as documented and a 14-day rolling memory. [Kiweely Vine](https://kiweely.com/vine/how-it-works) is a native macOS and Windows desktop agent that connects to your client’s tenants, reads configuration and data, and carries out approved changes across the UI, web services, Extend toolchain, and local Workday Studio.

| Dimension | Workday Deployment Agent (ASOR Ask) | Kiweely Vine Desktop Agent |
| --- | --- | --- |
|  **Supported Environments** | Implementation, Sandbox, Preview, Customer Central (US/EU regions; blocked in Production) | Personal developer tenants, Development, Implementation, Sandbox, and user-confirmed Production (with strict one-by-one read-back rules) |
|  **Tenant Configuration Writes** | None as documented (read-only inspection and text instructions); change-making is in partner early access via Configuration Cloud | Creates and updates calculated fields, custom reports, condition rules, business processes, and security with explicit consent |
|  **Data Loads (EIB)** | Explains SOAP/validation errors in chat | Generates templates, maps source data, validates transformations, and prepares inbound/outbound EIB configurations |
|  **Workday Extend & Orchestrate** | Not supported (separate Developer Agent scope) | Designs AMD/PMD/WQL/Orchestrations, validates schemas, deploys to Development tenants, and tests pages in a real browser |
|  **Workday Studio Integrations** | Interprets pasted SOAP error messages only | Drives your locally installed Eclipse-based Workday Studio via drop-in plugin to edit, build `.lar` files, deploy, launch, and inspect logs |
|  **Workflow & Persona Testing** | Manual verification by the admin after following steps | Visible, stoppable Chrome browser testing across personas, plus automated design workbooks and test scripts from live tenant state |
|  **Conversation & Project History** | Deleted permanently after 14 days; isolated per tenant | Durable, searchable conversations and deliverables organized by Client and Project on your machine |
|  **Commercial & Credit Model** | Requires UMSA; consumes Production Flex Credits in IMPL/Sandbox/Preview | Vine Free is free for your Workday personal developer account and its tenant. Any other tenant needs Vine Personal at $99 a month for one person, or a Vine Team seat at $149 a month. Paid seats include monthly Kiweely Credits or run on your own ChatGPT, Claude, or API plan |

## Where Consultants Lose Hours Between Diagnosis and Delivery

In real project delivery, knowing *which* task to open is rarely the bottleneck. Consultants lose hours in four places where read-only inspection cannot follow:

-   **Deep calculated field and condition rule chains:** A single benefits eligibility rule or composite report column often depends on five layered calculated fields across **Worker**, **Position**, and **Compensation**. [Auditing or building deep calculated field chains](https://kiweely.com/blog/auditing-deep-workday-calculated-field-chains) requires creating each dependent field in sequence, verifying return types, and testing edge cases against real worker records—not copying bullet points out of a chat pane.
-   **Multi-persona security and workflow verification:** Editing a domain policy or [business process condition rule](https://kiweely.com/guides/condition-rule-not-working) is half the job. Proving that a constrained **HR Partner** in EMEA sees the step while a US **Manager** does not requires structured [multi-persona proxy testing](https://kiweely.com/guides/workday-multi-persona-proxy-testing) in a real browser and capturing evidence before sign-off.
-   **Extend apps and Studio assemblies:** Neither Workday Extend apps nor Eclipse-based **Workday Studio** `.lar` integrations live inside Customer Central prompts. Kiweely Vine builds, validates, and deploys Extend apps end to end, and connects directly to your local Workday Studio installation (supporting Workday sign-in with saved details, Username and password, or OAuth API client with PKCE) to edit mediation routes, XSLT transforms, and integration systems.
-   **Cross-refresh continuity and client deliverables:** Because Workday purges Deployment Agent chats after 14 days and isolates them per tenant, your troubleshooting context vanishes right when a monthly Sandbox refresh wipes in-flight work. Kiweely Vine keeps every conversation, design workbook, and test script filed under its Client and Project so [rebuilding after a Sandbox refresh](https://kiweely.com/blog/rebuilding-after-workday-sandbox-refresh) starts from verified records rather than memory.

## How Consent-Gated Execution Works in Practice

Giving an AI agent write capability in Workday requires stricter boundaries than a read-only chat widget, not looser ones. In [Kiweely Vine](https://kiweely.com/vine), every conversation pins a single Client and only the specific tenants you tick for that thread; credentials for other clients or unselected tenants are never served to that turn.

Before changing anything in a tenant, the agent states what it plans to modify and waits for your approval. In non-Production Implementation or Sandbox tenants, you can review changes individually or authorize a specific list of named items—such as creating a batch of calculated fields—which the agent still executes and reads back one by one. If you explicitly connect a Production tenant, Full access is locked off by the host, bulk loads and EIB writes are barred, and every individual configuration change is verified before the next step proceeds.

Use Workday’s Deployment Agent when you want a quick, in-tenant citation from the Administrator Guide inside Customer Central. When you actually need the calculated fields built, the Studio `.lar` deployed, the Extend app validated, and the workflow tested in front of you, start in [Kiweely Vine](https://kiweely.com/vine/jobs).

## Frequently asked questions

### Can Workday's Deployment Agent create calculated fields or edit security policies?

As Workday documents it, no. Even with the Agent System of Record Ask skill and Tenant Inspection enabled, Workday's Deployment Agent is read-only. It inspects your tenant configuration and outputs step-by-step instructions that an administrator must enter manually in tasks such as Create Calculated Field or Edit Domain Security Policies. With Configuration Cloud, in opt-in early access for Workday partners since 18 Nov 2026, Deployment Agent can carry out changes once approved.

### Does Workday's Deployment Agent work in Production tenants?

No. You cannot access or query the Deployment Agent interactively in a Production environment. Workday recommends configuring its security policy and Agent Management Hub registration in Production solely so those settings copy down to Sandbox tenants during standard refreshes without requiring manual setup each week.

### Does the Deployment Agent consume Workday Flex Credits?

The basic Deployment Agent (Q&A) task in Customer Central is free and only searches Workday documentation and Community content. However, the ASOR-enabled Deployment Agent with the Ask skill is one of the few Workday-built agents whose usage in Implementation, Sandbox, and Preview tenants charges against your organization's Production Flex Credit balance, at 5 Flex Credits per Ask interaction in Implementation and Sandbox tenants.

### How long does Workday retain Deployment Agent chat history?

Workday isolates Deployment Agent chat history by tenant and permanently deletes conversations after 14 days. You cannot view a Sandbox conversation from a Preview tenant or preserve transcripts across a multi-month implementation inside the tool.

## Read next

[All posts](https://kiweely.com/blog)

-   [November 13 ### Why Flex Credit Forecasting Gets Harder When Sana, Data Cloud, and Custom Agents Share One Pool](https://kiweely.com/blog/estimating-workday-flex-credits-for-sana-and-data-cloud)
-   [November 13 ### Connecting Claude or Copilot to Workday via MCP: Seven Security Questions to Answer First](https://kiweely.com/blog/workday-mcp-agent-ready-tools-security-checklist)
-   [November 4 ### Using Glean with Workday Extend: what Workday documents and what it doesn’t](https://kiweely.com/blog/workday-extend-glean)

### More guides on Custom agents and Workday AI

[All 91 guides](https://kiweely.com/guides#custom-agents)

-   [Who gets Workday Developer Agent, and what it uses in Flex Credits](https://kiweely.com/guides/workday-developer-agent-access-flex-credits)
-   [Connecting a Workday custom agent to an outside agent platform with A2A](https://kiweely.com/guides/workday-custom-agents-a2a-external-agents)
-   [Why Workday custom agents use up Flex Credits so quickly](https://kiweely.com/guides/workday-custom-agent-usage-limits)
-   [A Workday custom agent’s lifecycle, from first deploy to Production](https://kiweely.com/guides/workday-custom-agent-lifecycle-promotion)
-   [Workday Illuminate, Agent System of Record, and Custom Agents: What Each Covers](https://kiweely.com/guides/workday-illuminate-asor-custom-agents-comparison)
-   [Can Workday Extend build custom agents? What you can and cannot build](https://kiweely.com/guides/workday-extend-custom-agents)
-   [How to connect your own agent to Workday’s Agent System of Record, step by step](https://kiweely.com/guides/connect-custom-agent-workday-system-of-record)
-   [Workday Agent System of Record (ASOR): An Admin’s Security and Setup Checklist](https://kiweely.com/guides/workday-agent-system-of-record-security)

## Your Workday work, done by AI.

Kiweely Vine is the AI agent that builds, deploys and tests Extend apps as real people, makes changes in any Workday task screen, answers from your real tenant and builds reports.

[Start for free](https://kiweely.com/download) [See how it works](https://kiweely.com/vine/how-it-works)

## Get the next one in your inbox.

One email a month for Workday teams: what’s changing in Workday and what to retest, one tip tried in a real tenant, and what’s new at Kiweely.
