richy.mulcahy

Since this is a Workato Owned Community connector which you can see here. if there is a bug you can submit a support ticket for our team to look at

This is a community connector created by our team. it was originally created for a specific use case and then published for others to use as well. you can see a few more details following the link here app.workato.com/browse/connectors/1092945/details. While the documentation on community connectors is less than Platform connectors the plus side is you have access to the code. So you can put that into claude or any other agent for analysis or modifications.

agree the wording of manual entry and adding the .txt requires a couple reads

when you say REST API do we already have a connector? in my screenshot using google drive to download file then map that datapill to the jira file contents. This is also possible with a HTTP connection to an API we dont have a connector for but is a bit more complex since youll need to configure the action with the correct type (GET, POST) and the response body structure for the file contents return

yea as long as you dont need cycles with the LLM MCP app should work.

from my understanding this is the high level differences MCP apps live inside the tool itself so any AI client can render them (Claude, ChatGPT, etc), and once launched the app calls other tools directly without going back to the AI. Great for reviewing/approving data that already exists. also part of the protocol so not tied to a vendor. Artifacts stay in conversation with the AI the whole time, so they're the better fit when you want to keep revising content back and forth. but i believe tied to claude. Quick question for you, once you see an email draft, are you usually asking the AI to tweak or rewrite it, or mostly just reviewing and sending as is? if your looking to keep conversing with the LLM then and atifact is likely better. if its more review, make a change, and then approve/reject MCP app is great. The MCP app will also likely have lower token costs since its not involving the model every step

Shivam are you talking about this feature? docs.workato.com/en/mcp/mcp-apps

RussellJ very well said and even dropped one of my favorite sayings ever. Although i go with the version of "to a hammer everything is a nail" John A. the broad scope of Netsuite can be difficult to wrap your arms around. along with all the things mentioned by Russell I like to try and simplify the integration possibilities by breaking it down to just 2 applications in scope (one being Netsuite). So when talking to your team what are the most common other applications that work with Netsuite processes? then you can begin building the bridge between them. I have included some of our more popular Netsuite Accounting related use cases below. Plus we are seeing large demand for Netsuite MCPs to expose this information to finance users directly in their Agent of choice.

- Journal entry automation — pull data from a source system (Snowflake, data warehouse) and push JEs into NetSuite on a schedule, replacing manual file handoffs and cutting a 45 min process down to under 20 min for high volume batches
- Invoice automation — connect a billing or platform tool directly to NetSuite to eliminate manual journal entry creation, then expand into other tool to NetSuite invoice automation
- Order to cash — CRM to NetSuite integration for creating new records, syncing quotes and orders through to invoicing
- Cross tool data sync — sync a project or planning tool (like Smartsheet) with NetSuite for financial tracking
- Approval notifications — push NetSuite journal entry or transaction approvals into Slack or Teams so approvers don't need to log into NetSuite directly
- Large volume AR reconciliation — for accounts with 10k+ open invoices, use a data pipeline approach (JDBC/SuiteAnalytics Connect) instead of standard API recipes, since NetSuite's REST/SOAP layers hit row and timeout limits at scale
- AP automation — email inbox to PDF extraction to PO validation to bill creation, fully automated on matched invoices with exception routing for anything that doesn't match

Nikhil K. i chatted a bit with some of our workday experts and with the non batch trigger i do believe the recipe job queue would max and out likely fail. I believe the full process is we send the Get request for all new/updated workers, Workday sends back that there are 150k and on the first page gives the first 999. we then turn those into 999 jobs and ask for page 2 repeating through the 150k. however the polling queue will will at 10k causing errors. This is where batch would work. batching 50 at a time would put the cap of the jobs at 3k. so while it would take a while to process the queue would not fill. youd want to process each back in a loop in the main recipe since dumping them to a recipe function would also fill that queue.

working to confirm but batching may be required. polling triggers typlically have a max queue of 10k jobs