MULTI-APP TOOLKIT
Build with Google Gemini + Jira Service Desk
- Multi-app
- One MCP session
- 23 actions
- 2 triggers
- Managed auth
MCP
One MCP session, both toolsets
One session gives your product or agent every Google Gemini and Jira Service Desk tool at once — connect once, list tools, and call them like any single-app session.
// accessToken: mint a short-lived token with the Connect SDK — see the MCP guide
const transport = new StreamableHTTPClientTransport(
new URL("https://remote.mcp.pipedream.net/v3"),
{
requestInit: {
headers: {
Authorization: `Bearer ${accessToken}`,
"x-pd-project-id": "{project_id}",
"x-pd-environment": "production",
"x-pd-external-user-id": "{external_user_id}", // any stable ID for this user in your system
"x-pd-app-slug": "google_gemini,jira_service_desk",
},
},
},
)
const mcp = new Client({ name: "my-agent", version: "1.0.0" })
await mcp.connect(transport)
const { tools } = await mcp.listTools()
// One list, both toolsets: Google Gemini and Jira Service Desk tools arrive
// together, each keyed by its own app's slug.
// e.g. run Generate Content from Text:
const result = await mcp.callTool({
name: "google_gemini-generate-content-from-text",
arguments: {
model: "Model",
text: "Prompt Text",
},
})import { PipedreamClient } from "@pipedream/sdk"
const pd = new PipedreamClient({
projectId: process.env.PIPEDREAM_PROJECT_ID!,
clientId: process.env.PIPEDREAM_CLIENT_ID!,
clientSecret: process.env.PIPEDREAM_CLIENT_SECRET!,
projectEnvironment: "production",
})
const externalUserId = "{external_user_id}" // any stable ID for this user in your system
const [googleGeminiTools, jiraServiceDeskTools] =
await Promise.all([
pd.components.list({ app: "google_gemini" }),
pd.components.list({ app: "jira_service_desk" }),
])
// One external user owns both connected accounts, so either app's tools
// run on their behalf with the same externalUserId.curl "https://api.pipedream.com/v1/connect/{project_id}/components?app=google_gemini" \
-H "X-PD-Environment: production" \
-H "Authorization: Bearer {access_token}"
curl "https://api.pipedream.com/v1/connect/{project_id}/components?app=jira_service_desk" \
-H "X-PD-Environment: production" \
-H "Authorization: Bearer {access_token}"
# Connect both accounts to the same external_user_id, then
# configure and invoke either app's tools on that user's behalf.ARCHITECTURE
One user, two connected accounts
Your user connects each account once, under whatever ID they already have in your product. The two stay independent — either can be revoked on its own — and your code reaches both through that one user.
Your product
external_user_id
{external_user_id}Pipedream Connect
Managed identity
Auth · tools · routing
Google Gemini
Connected account
Jira Service Desk
Connected account
TOOLS
Google Gemini tools
The Google Gemini tools this pairing puts in reach, each one running on your user's own connected account.
Actions
-
Generate Content from Text
actionGenerates content from text input using the Google Gemini API. See the documentationWritev0.2.2 -
Generate Content from Text and Image
actionGenerates content from both text and image input using the Gemini API. See the documentationWritev1.0.3 -
Generate Embeddings
actionGenerate embeddings from text input using Google Gemini. See the documentationWritev0.0.2
Triggers
No Google Gemini triggers are available yet.
TOOLS
Jira Service Desk tools
The Jira Service Desk tools this pairing puts in reach, each one running on your user's own connected account.
Actions top 8 of 20
-
Create Comment on Request
actionCreate a comment on a customer request. See the documentationWritev0.1.5 -
Create Incident
actionCreates a new incident request in a Jira Service Desk. Auto-discovers the best-matching request type (prefers types named 'incident' or 'problem') and, ifserviceDeskIdis omitted, auto-discovers it from the user's existing requests. Use List Sites first to obtain the requiredcloudId.serviceDeskIdis optional — if not provided, the tool calls List My Requests to discover it automatically. Returns the new request including itsissueKeyandissueId. See the documentationWritev0.0.2 -
Create Request
actionCreates a customer request (ticket) in a Jira Service Management service desk. This is the single tool for creating any kind of ticket (incident, service request, access request, hardware request, and so on). The kind of ticket is decided byrequestTypeId, not by the wording of the summary, so always pick the request type deliberately. Use List Sites to getcloudId, List Service Desks to getserviceDeskId, and List Request Types to choose therequestTypeIdwhose name and description match the user's intent. Call List Request Type Fields to see which fields that request type requires; pass anything beyond summary and description inadditionalFieldValues, keyed by Jira field ID. Worked example: on service desk1, request type4("Onboard new employees") requiressummaryand also accepts aduedate, so call with SummaryJoseph Wilson starts on September 1, DescriptionNeeds a laptop and an email account, and Additional Field Values{ "duedate": "2026-09-01" }. Optionally attach one or more files at creation time viaattachments; to add, replace, or delete attachments on a request that already exists, use Manage Request Attachment instead. Returns the created request including itsissueKeyandissueId. Ifattachmentsis set, the response also includes either anattachmentsarray (on success) or anattachmentErrorstring (if the request was created but the attachment step failed) — the request itself is never rolled back because of an attachment failure. See the documentationWritev1.1.1 -
Download Issue Attachment
actionDownload the binary content of a Jira Service Desk attachment to the file-stash directory, returning the saved path plus the attachment metadata (filename,mimeType,size). Run List Issue Attachments first to obtain the attachmentid, then pass both it and the sameissueIdOrKeyhere. Example: passingissueIdOrKeyIT-42andattachmentId10042downloadsscreenshot.pngto/tmp/10042-screenshot.pngand returns{ "filedata": ["10042-screenshot.png", "/tmp/10042-screenshot.png"], "attachment": { "id": "10042", "filename": "screenshot.png", "mimeType": "image/png", "size": 84213 } }. See the documentationRead-onlyv0.0.4 -
Find Service Desk Customers
actionFinds the customers of one service desk by name or email address and returns theaccountIdof each match. Pick this tool when the person is the one the ticket is being raised for on a service desk you can identify, i.e. to fillraiseOnBehalfOfon Create Request. Natural-language cues: "raise a ticket for Jean on the IT desk", "open a request on behalf of john@acme.com", "file this for my colleague Dana", "submit a hardware request for the new starter". Pick Find Users instead when the person does not have to be a customer of this desk, such as an approver, manager, or agent you are adding torequestParticipants, or when you cannot tell which service desk applies. Prefer this tool wherever both would work: it searches only this desk's customer list, so a match proves the person can actually raise a request here, and bot accounts are excluded (a site-wide search on a live site returned 17 users of which 16 were integrations). Use List Sites forcloudIdand List Service Desks forserviceDeskIdfirst. Worked example: for "open a laptop request for Joseph Wilson on the IT desk", call this with Service Desk ID1and QueryJoseph Wilson, readaccountId5b10a2844c20165700ede21goff the single match, then call Create Request with Service Desk ID1and thataccountIdasraiseOnBehalfOf. Omit Query to list every customer of the desk, which answers "who can raise requests on this desk?". Query is matched againstdisplayNameandemailAddress, and matches more than just the start of them. Pass a full name or a full email address to keep the result set tight. If nobody matches, the person may exist on the site without being a customer of this desk, retry with Find Users. Results are paginated automatically up tomaxResults. Returns{ users, truncated }, wheretruncatedistruewhen more matches remained unfetched.accountIdis the only field guaranteed present: Atlassian's profile visibility rules hideemailAddresson users who have not made it public, so match ondisplayNameand never require an email to be returned. An unknown or inaccessible Service Desk ID fails with a 404 rather than returning an empty list, so an empty list really does mean nobody matched. See the documentationRead-onlyv0.0.1 -
Find Users
actionFinds any active user on an Atlassian site by name or email address and returns theaccountIdof each match. Pick this tool when the person does not have to be a customer of one particular service desk, i.e. to fillrequestParticipantson Create Request with an approver, manager, watcher, or agent, or when the user names somebody but no service desk is known yet. Natural-language cues: "add my manager Dana as a participant", "cc the security lead on this ticket", "loop in john@acme.com", "what is the account ID for Jean?". Pick Find Service Desk Customers instead when you already know the service desk and the person is the one the ticket is being raised for (raiseOnBehalfOf). That tool is more precise, confirms the person can actually raise a request on that desk, and excludes bots. Use List Sites first to obtain the requiredcloudId. Worked example: for "open a laptop request and add Dana Lee as a participant", call this with QueryDana Lee, readaccountId5b10a2844c20165700ede21goff the match whoseaccountTypeisatlassian, then pass["5b10a2844c20165700ede21g"]asrequestParticipantson Create Request. Most users on a Jira site are bots, not people. Integrations come back as ordinary matches carryingaccountTypeapp, so readaccountTypeand use onlyatlassianaccounts asraiseOnBehalfOforrequestParticipants. Query is matched againstdisplayNameandemailAddress, and matches more than just the start of them. Pass a full name or a full email address to keep the result set tight. Results are paginated automatically up tomaxResults. Returns{ users, truncated }.truncatedistruewhen the result set may be incomplete, either because more matches remained unfetched or because collection stopped at Atlassian's 1000-match limit, which looks the same whether or not further matches exist. Once you have 1000 matches, narrow the query rather than raisingmaxResults, which cannot go higher.accountIdis the only field guaranteed present: Atlassian's profile visibility rules hideemailAddresson users who have not made it public, so match ondisplayNameand never require an email to be returned. An emptyuserslist means either nobody matched or the connected account lacks the "Browse users and groups" global permission, which Atlassian reports as zero results rather than as an error. See the documentationRead-onlyv0.0.1 -
Get Current User
actionReturns the authenticated user'saccount_id,display_name, andemailfrom Atlassian. Use this to identify who is logged in, or to filter requests by the current user'saccount_id. NocloudIdrequired — this uses the Atlassian Identity API directly. See the documentationRead-onlyv0.0.8 -
Get Request
actionFetches the full details of a Jira Service Desk request including its field values and comment thread in a single response. Comments are paginated automatically up tomaxResults; the summary says so when the thread was truncated. Use this to summarize a ticket without needing follow-up calls. Use List Sites first to obtain the requiredcloudId. Use List My Requests to find theissueKeyof a request (e.g.IT-42). See the documentationRead-onlyv0.2.5
MULTI-APP
Works with more apps
The combinations customers connect alongside these two — nothing here is limited to a pair.
REFERENCE
Toolkit details
- x-pd-app-slug
- google_gemini,jira_service_desk
- Primary app
- Google Gemini (google_gemini)
- Second app
- Jira Service Desk (jira_service_desk)
- Authentication
- API key + OAuth
- Available actions
- 23
- Available triggers
- 2