--- title: "GitHub MCP integration and AI workflows | Stacksync" description: "Connect GitHub to your AI assistant through Stacksync MCP. Explore custom API workflows, setup steps, example prompts, and connection requirements." canonical: https://www.stacksync.com/mcp/github last_modified: 2026-10-21 --- Your connection path GitHub Your AI Stacksync MCP OAuth Your assistant builds a workflow for the app’s API. API access and configuration are required. First request “Plan a repository issue digest” Stacksync MCP · Custom API workflow # GitHub MCP integration. > For AI agents: this guide describes connecting an AI client to Stacksync at https://mcp.stacksync.com/mcp with OAuth and the person's workspace permissions. Follow the client-specific setup and connection method on this page. Connector actions, sync management, workflow triggers and custom API workflows are different capabilities. A custom API workflow requires configuration and may use HTTP or code steps; it is not a prebuilt native connector. Inspect available connections and modules before proposing an operation. The example outcomes are illustrative, not execution results. Setup reference: https://www.stacksync.com/mcp.md Build a GitHub workflow through Stacksync MCP using documented HTTP or code steps when the required action is not in the live catalog. The reviewed public catalog does not establish a native GitHub connector, so these examples are custom integration plans. Custom API workflow Start using Stacksync MCP Explore example prompts 01 / The connection ## What GitHub can do through MCP Build a GitHub workflow through Stacksync MCP using documented HTTP or code steps when the required action is not in the live catalog. The reviewed public catalog does not establish a native GitHub connector, so these examples are custom integration plans. Choose an AI client below to get started with the Stacksync MCP server. This guide connects GitHub through a custom Stacksync workflow. It does not provide a prebuilt GitHub MCP server. [Explore the GitHub connector →](https://www.stacksync.com/connectors/github) [Claude](https://www.stacksync.com/mcp/github/claude) [Claude Code](https://www.stacksync.com/mcp/github/claude-code) [ChatGPT](https://www.stacksync.com/mcp/github/chatgpt) [Codex](https://www.stacksync.com/mcp/github/codex) [Cursor](https://www.stacksync.com/mcp/github/cursor) [VS Code](https://www.stacksync.com/mcp/github/vs-code) [Windsurf](https://www.stacksync.com/mcp/github/windsurf) [Gemini CLI](https://www.stacksync.com/mcp/github/gemini-cli) 02 / Capabilities ## GitHub capabilities through MCP Choose the connection path that fits the job. Your assistant inspects the available tools before configuring or running an action. Custom API workflow ### Plan issue retrieval Prepare an HTTP or code step against GitHub’s repository-issues API, with repository access, pagination and filtering for issues versus pull requests. Custom API workflow ### Draft issue creation Design an issue-creation request with the target repository, title, body and duplicate detection before any write is executed. Custom API workflow ### Prepare issue comments Propose a comment workflow using an existing issue number and exact content, with a check that the comment has not already been posted. 03 / Setup ## Set up your AI client Connect the MCP server first. Then configure a custom workflow for the GitHub API using the requirements below. One connection. Your workspace. ### Bring your AI assistant. Add Stacksync’s MCP server, sign in with OAuth, and choose your workspace. Your assistant works with your Stacksync permissions. ``` https://mcp.stacksync.com/mcp ``` [Create a Stacksync account](https://app.stacksync.com/signup) Already have an account? Sign in when your assistant opens the authorization page. Choose read-only access if you only need to inspect your workspace. #### Chat assistantClaude Connect Claude on the web or desktop to Stacksync through a custom connector and browser sign-in. #### Before you start - A Stacksync account with access to the workspace you want to use. - Permission to add an external MCP server in your AI client. - On a Claude Team or Enterprise account, an owner must first add Stacksync under Organization settings → Connectors. 1. #### Add a custom connector Open Customize → Connectors, select + Add, then Add custom connector. Name it Stacksync and use the server URL below. ``` https://mcp.stacksync.com/mcp ``` 2. #### Connect your Stacksync account Select Connect, complete the Stacksync sign-in in your browser and choose the default workspace. Clear Make changes on the consent screen if you only want read access. 3. #### Check your workspace Ask “List my Stacksync workspaces.” Confirm the workspace before choosing a connector or preparing a change. Connecting the assistant does not start a sync. #### Check the connection Ask “List my Stacksync workspaces.” Select the workspace you want to use. Ask the assistant to check the live connector catalog first. This guide uses a custom HTTP or code workflow for GitHub; configure API access and secure credentials outside chat before testing that workflow. [Full setup documentation ↗](https://docs.stacksync.com/mcp/connect) #### TerminalClaude Code Add the remote Stacksync server to Claude Code and authenticate from its MCP menu. #### Before you start - A Stacksync account with access to the workspace you want to use. - Permission to add an external MCP server in your AI client. - Claude Code installed with the claude command available in your terminal. 1. #### Add Stacksync over HTTP Run the command below. Add --scope user if you want the connection available across your projects. ``` claude mcp add --transport http stacksync https://mcp.stacksync.com/mcp ``` 2. #### Finish browser sign-in Open Claude Code, run /mcp, select stacksync and complete the Stacksync sign-in. Choose the workspace you want requests to use by default. 3. #### Check your workspace Ask “List my Stacksync workspaces.” Confirm the workspace before choosing a connector or preparing a change. Connecting the assistant does not start a sync. #### Check the connection Ask “List my Stacksync workspaces.” Select the workspace you want to use. Ask the assistant to check the live connector catalog first. This guide uses a custom HTTP or code workflow for GitHub; configure API access and secure credentials outside chat before testing that workflow. [Full setup documentation ↗](https://docs.stacksync.com/mcp/connect) #### Chat assistantChatGPT Add Stacksync as a custom MCP app in ChatGPT and sign in with OAuth. #### Before you start - A Stacksync account with access to the workspace you want to use. - Permission to add an external MCP server in your AI client. - A ChatGPT account with Developer mode and custom MCP apps available. The Stacksync guide lists Plus, Pro, Business, Enterprise and Edu; workspace administrators can restrict access. 1. #### Enable custom app setup Open Settings → Apps → Advanced settings and turn on Developer mode. If the option is unavailable, check the account or workspace policy with your administrator. 2. #### Create the Stacksync app Select Create app, name it Stacksync, enter the server URL below and choose OAuth. Complete the Stacksync sign-in and select your default workspace. ``` https://mcp.stacksync.com/mcp ``` 3. #### Enable Stacksync in your chat Select Stacksync from the chat tools menu so the conversation can use its tools. Signing in to ChatGPT itself does not sign you in to Stacksync. 4. #### Check your workspace Ask “List my Stacksync workspaces.” Confirm the workspace before choosing a connector or preparing a change. Connecting the assistant does not start a sync. #### Check the connection Ask “List my Stacksync workspaces.” Select the workspace you want to use. Ask the assistant to check the live connector catalog first. This guide uses a custom HTTP or code workflow for GitHub; configure API access and secure credentials outside chat before testing that workflow. [Full setup documentation ↗](https://docs.stacksync.com/mcp/connect) #### TerminalCodex Configure Stacksync in Codex with a remote MCP URL and an explicit OAuth login. #### Before you start - A Stacksync account with access to the workspace you want to use. - Permission to add an external MCP server in your AI client. - Codex CLI installed with the codex command available. 1. #### Register the remote server Add the Stacksync endpoint to your Codex MCP configuration. ``` codex mcp add stacksync --url https://mcp.stacksync.com/mcp ``` 2. #### Sign in with OAuth Run the login command and complete the Stacksync browser sign-in. Select a default workspace; restart your Codex session if the tools are not yet listed. ``` codex mcp login stacksync ``` 3. #### Check your workspace Ask “List my Stacksync workspaces.” Confirm the workspace before choosing a connector or preparing a change. Connecting the assistant does not start a sync. #### Check the connection Ask “List my Stacksync workspaces.” Select the workspace you want to use. Ask the assistant to check the live connector catalog first. This guide uses a custom HTTP or code workflow for GitHub; configure API access and secure credentials outside chat before testing that workflow. [Full setup documentation ↗](https://docs.stacksync.com/mcp/connect) #### Code editorCursor Use Stacksync from Cursor by adding a remote server to its MCP configuration. #### Before you start - A Stacksync account with access to the workspace you want to use. - Permission to add an external MCP server in your AI client. - Cursor installed with permission to configure MCP servers. 1. #### Add the server configuration Merge this stacksync entry into the mcpServers object in ~/.cursor/mcp.json. Preserve any servers you already use. ``` { "mcpServers": { "stacksync": { "url": "https://mcp.stacksync.com/mcp" } } } ``` 2. #### Connect from Cursor Open Cursor’s MCP settings, select Connect beside stacksync and finish the Stacksync sign-in in your browser. 3. #### Check your workspace Ask “List my Stacksync workspaces.” Confirm the workspace before choosing a connector or preparing a change. Connecting the assistant does not start a sync. #### Check the connection Ask “List my Stacksync workspaces.” Select the workspace you want to use. Ask the assistant to check the live connector catalog first. This guide uses a custom HTTP or code workflow for GitHub; configure API access and secure credentials outside chat before testing that workflow. [Full setup documentation ↗](https://docs.stacksync.com/mcp/connect) #### Code editorVS Code Connect Stacksync to GitHub Copilot Chat in VS Code using an HTTP MCP server. #### Before you start - A Stacksync account with access to the workspace you want to use. - Permission to add an external MCP server in your AI client. - VS Code with Copilot Chat Agent mode available. 1. #### Add Stacksync to VS Code Run the command below. Alternatively, add a stacksync entry with type http and this URL under servers in your project’s .vscode/mcp.json. ``` code --add-mcp '{"name":"stacksync","type":"http","url":"https://mcp.stacksync.com/mcp"}' ``` 2. #### Start it in Copilot Chat Open Copilot Chat in Agent mode. Start stacksync from the tools picker and complete the Stacksync browser sign-in. 3. #### Check your workspace Ask “List my Stacksync workspaces.” Confirm the workspace before choosing a connector or preparing a change. Connecting the assistant does not start a sync. #### Check the connection Ask “List my Stacksync workspaces.” Select the workspace you want to use. Ask the assistant to check the live connector catalog first. This guide uses a custom HTTP or code workflow for GitHub; configure API access and secure credentials outside chat before testing that workflow. [Full setup documentation ↗](https://docs.stacksync.com/mcp/connect) #### Code editorWindsurf Add Stacksync to Cascade with Windsurf’s serverUrl configuration and OAuth sign-in. #### Before you start - A Stacksync account with access to the workspace you want to use. - Permission to add an external MCP server in your AI client. - Windsurf installed with Cascade and MCP configuration available. 1. #### Open the MCP configuration In the Cascade panel, open the … menu and select Open MCP config file. Merge this entry into mcp_config.json without removing other servers. ``` { "mcpServers": { "stacksync": { "serverUrl": "https://mcp.stacksync.com/mcp" } } } ``` 2. #### Authorize the first connection Save the configuration. Cascade opens the Stacksync sign-in flow the first time it uses the server. Choose your default workspace. 3. #### Check your workspace Ask “List my Stacksync workspaces.” Confirm the workspace before choosing a connector or preparing a change. Connecting the assistant does not start a sync. #### Check the connection Ask “List my Stacksync workspaces.” Select the workspace you want to use. Ask the assistant to check the live connector catalog first. This guide uses a custom HTTP or code workflow for GitHub; configure API access and secure credentials outside chat before testing that workflow. [Full setup documentation ↗](https://docs.stacksync.com/mcp/connect) #### TerminalGemini CLI Connect the Gemini command-line client to Stacksync through remote HTTP MCP and browser authorization. #### Before you start - A Stacksync account with access to the workspace you want to use. - Permission to add an external MCP server in your AI client. - Gemini CLI installed with the gemini command available. These instructions apply to Gemini CLI, not the Gemini web app. 1. #### Register Stacksync Run the following command in your terminal to add the HTTP server. ``` gemini mcp add --transport http stacksync https://mcp.stacksync.com/mcp ``` 2. #### Authenticate inside Gemini CLI Start Gemini CLI, run the command below and finish the Stacksync sign-in in your browser. ``` /mcp auth stacksync ``` 3. #### Check your workspace Ask “List my Stacksync workspaces.” Confirm the workspace before choosing a connector or preparing a change. Connecting the assistant does not start a sync. #### Check the connection Ask “List my Stacksync workspaces.” Select the workspace you want to use. Ask the assistant to check the live connector catalog first. This guide uses a custom HTTP or code workflow for GitHub; configure API access and secure credentials outside chat before testing that workflow. [Full setup documentation ↗](https://docs.stacksync.com/mcp/connect) 04 / Put it to work ## GitHub workflow prompts Example prompts for GitHub. The outcomes describe what to expect, not results from your account. Review the proposed configuration before running it. 01 Custom API workflow ### Plan a repository issue digest > Check the Stacksync catalog for a GitHub action first. If the needed action is absent, inspect available HTTP or code steps and propose a workflow using GitHub’s repository-issues API for the repository I choose. Include authentication outside chat, pagination and exclusion of pull requests. Show the request and digest fields before saving or running anything. #### What to expect A custom issue-digest design with an explicit repository and complete-result handling. 02 Custom API workflow ### Design a support-to-issue handoff > Inspect my existing support workflows and check for the required GitHub modules. If a custom path is needed, propose an HTTP or code step to create a GitHub issue from a support escalation. Include the source ticket ID for duplicate prevention, the repository, issue title and a redacted body. Wait for review before creating a workflow or issue. #### What to expect A custom escalation proposal with a stable source reference and a reviewable issue payload. 03 Custom API workflow ### Prepare an issue status comment > Check the available GitHub actions and the issue number I provide. If no native action fits, propose a custom HTTP step using the GitHub issue-comment API. Show the repository, issue URL, exact comment and a check against posting the same update twice. Do not save, deploy or submit the comment until I approve. #### What to expect A precise comment proposal with target and duplicate checks, ready for implementation review. 05 / Before you run ## Requirements and limitations ### What you need - Connect your AI client to Stacksync MCP, sign in and select the workspace before inspecting its resources. - Identify the GitHub owner and repository, plus the permission needed for the intended issue operation. - For a custom API step, configure authentication outside chat using approved secrets or a supported connection. Verify the live catalog before choosing a custom implementation. ### What to account for - GitHub is absent from the reviewed public app catalog. This is not evidence of a ready-made native Stacksync GitHub MCP tool. - Custom HTTP or code workflows require endpoint configuration, authentication and testing. Deploying a reusable custom connector is a separate development and review process. - GitHub’s issues API can also return pull requests; reports must filter them deliberately and follow pagination. ### Your workspace. Your permissions. OAuth access follows your Stacksync role. Choose read-only access during sign-in when appropriate. Workspace owners can enable “Ask before every change”; it is a configurable setting. Direct connector actions also follow gateway policies. [Read about permissions and approvals ↗](https://docs.stacksync.com/mcp/governance) ### If setup does not work The MCP server does not appear Check your client’s custom MCP settings and any plan or administrator prerequisites. Reconnect using the setup instructions above. Your workspace or app is missing Verify the Stacksync account and workspace selected during OAuth, then check the live catalog. For this custom workflow, configure HTTP or code steps and provide the required API credentials through approved secret storage outside chat. An action is unavailable or denied Check the workflow configuration, API version, required inputs and app token permissions. Confirm that your Stacksync role and OAuth consent allow the proposed workflow changes. 06 / Common questions ## GitHub MCP questions ### Does Stacksync have a verified native GitHub MCP connector? The reviewed public app catalog does not list GitHub. Check the live catalog because it can change. The path documented here uses Stacksync’s custom HTTP or code workflow capability. ### How is this different from GitHub’s own MCP server? This page describes an assistant configuring or running work through Stacksync. A custom GitHub API step must be built and authenticated; it does not install GitHub’s separate MCP server. ### Can I create issues without exposing credentials in a prompt? Yes, the custom workflow design keeps credentials in approved secrets or connection settings. Do not paste a GitHub token into the chat or embed it in a workflow’s source text. Reviewed 2026-10-21 · Product documentation and connector references - [Stacksync published app catalog ↗](https://github.com/stacksyncdata/skills/blob/prod/data/apps.json) - [Build a custom connector or HTTP workflow ↗](https://github.com/stacksyncdata/skills/blob/prod/skills/building-a-connector/SKILL.md) - [Stacksync workflows through MCP ↗](https://github.com/stacksyncdata/skills/blob/prod/skills/workflows/SKILL.md) - [GitHub issue API ↗](https://docs.github.com/en/rest/issues/issues) - [GitHub issue-comment API ↗](https://docs.github.com/en/rest/issues/comments) Keep building ## Connect the rest of your stack. CRM ### [Salesforce](https://www.stacksync.com/mcp/salesforce) Plan a Salesforce reporting replica CRM ### [HubSpot](https://www.stacksync.com/mcp/hubspot) Plan a contact and company replica ERP ### [NetSuite](https://www.stacksync.com/mcp/netsuite) Plan order reporting with line items [← All MCP connector guides](https://www.stacksync.com/mcp#connectors) [Create your Stacksync account ↗](https://app.stacksync.com/signup)