Important
This repository is under development. Additional connector-family skills (connect-ai-crm, connect-ai-erp, connect-ai-hcm, connect-ai-ticketing, connect-ai-analytics, connect-ai-files) and connector-specific skills are in development.
Note
connect-ai-base must be enabled alongside any other skill in this repo — all connector-family and connector-specific skills build on top of it. Always include --skill connect-ai-base when installing other skills individually.
Install every skill in this repo:
npx skills add CDataSoftware/connect-ai-skills
Install a specific skill by name (the folder name under skills/). Pair it with connect-ai-base:
npx skills add CDataSoftware/connect-ai-skills --skill connect-ai-base --skill connect-ai-salesforce
Pass --skill multiple times to install several skills in one command:
npx skills add CDataSoftware/connect-ai-skills --skill connect-ai-base --skill connect-ai-salesforce --skill connect-ai-jira
| Skill | Family | Description |
|---|---|---|
connect-ai-base |
— | Base skill. Required connection confirmation, discovery workflow for the generic MCP Server and Toolkits, query construction, and common error recovery. Load this first — all other skills compose on top of it. |
connect-ai-salesforce |
CRM | Salesforce data model, query patterns, stored procedures, and Salesforce-specific conventions. |
connect-ai-bullhorncrm |
CRM | Bullhorn CRM recruiting data model (Candidates, ClientCorporations, JobOrders, JobSubmissions, Placements), file-attachment procedures, edit-history tables, and Bullhorn conventions. |
connect-ai-workday |
HCM | Workday's multi-connection-type model (REST/WQL/Reports/SOAP), per-connection-type references, prompt-column lookups, value tables, and change-resource procedures. |
connect-ai-quickbooksonline |
Accounting | QuickBooks Online data model (GL, AP, AR, items, banking), line-item patterns, LineAggregate XML for inserts, and report objects. |
connect-ai-sageintacct |
Accounting | Sage Intacct's dual XML/REST surfaces, routed via getSchemas, with per-surface data models, the UseLegacy performance note, stored procedures, and write operations. |
connect-ai-stripe |
Accounting | Stripe's dual classic/V2 surfaces, routed via getSchemas, with the customer-centric billing/payments model, lifecycle stored procedures, cloud-compatible file download patterns, and Stripe-specific conventions. |
connect-ai-jira |
Ticketing | Jira data model, issue hierarchy, query patterns, stored procedures, and Jira-specific conventions. |
connect-ai-confluence |
Collaboration | Confluence data model (spaces, pages, comments, attachments, users, and analytics/hierarchy views), the space → page → refine workflow, and base64 image/attachment retrieval. |
connect-ai-airtable |
Collaboration | Airtable's multi-schema model (Information metadata plus one schema per base), per-base table/view anatomy, attachment/collaborator fields, comments, and stored procedures. |
connect-ai-docusign |
Collaboration | DocuSign data model (Envelopes, Templates, Documents, Recipients), envelope lifecycle queries, and procedures for creating/downloading documents. |
connect-ai-googlecalendar |
Collaboration | Google Calendar data model (events, calendars, ACLs, attachments), the per-calendar table pattern, availability queries, and event create/move procedures. |
connect-ai-googlesheets |
Collaboration | Google Sheets data model (Spreadsheets and Sheets metadata, dynamic per-sheet data tables), the spreadsheet→sheet→data query flow, the create-spreadsheet/add-sheet/insert workflow, and stored procedures. |
connect-ai-googledrive |
Files | Google Drive data model (files, folders, drives, permissions) and procedures for uploading and downloading file content. |
Additional skills will be listed here as they ship.
- CData Connect AI account with at least one connection configured
- The Connect AI MCP server added to your AI integration (see Connect AI Integrations - AI Tools)