# Stax.ai MCP connector The Stax.ai connector brings your stacks, documents, and TPA plans to Claude, ChatGPT, Cursor, and the Elaichi Agent, so your team can find, file, and organize accounting paperwork in plain language while every action is logged. Source: https://elaichi.ai/connectors/staxai/ ## Facts | | | | --- | --- | | Application | Stax.ai | | Category | Accounting | | AI tools | 32 | | Authentication | Connects with an API key | | Needs your own OAuth app | No | | MCP endpoint | https://api.elaichi.ai/mcp | | Works with | Claude, ChatGPT, Cursor, any MCP client, and the Elaichi Agent | | Tools advertised by name | No. 32 tools is past the 30-tool threshold, so clients use `search_tools` and `execute_tool` | ## What you can ask once Stax.ai is connected - List the stacks with documents added this week. - How many documents sit in the Payables stack? - Move last month's invoices into the Archive stack. ## Connect Stax.ai in Elaichi This happens once for the organization, before any client is involved. 1. Open Connections, choose Add connection, and pick Stax.ai. 2. Optionally set Share with, then press Connect. 3. Paste a Stax.ai API key. One person generates a token in Stax.ai and pastes it once. Everyone else works through Share with, and never sees it. Credentials are vaulted and nobody, including the AI, reads them back. The connection becomes a toolbox immediately, so you can curate which Stax.ai tools are exposed, rename them, or freeze arguments before anyone points a client at it. ## Connect Stax.ai to Claude Endpoint: https://api.elaichi.ai/mcp 1. Open Customize, then Connectors. 2. Press Add. 3. Name it, paste the MCP server URL, then Continue. 4. Sign in and approve. On Team and Enterprise, an Owner adds it once. Everyone else turns it on for themselves. ## Connect Stax.ai to ChatGPT Endpoint: https://api.elaichi.ai/mcp 1. Open Plugins, then press the + button. 2. Name it and paste the endpoint into Server URL. 3. Leave Authentication on OAuth, then tick the risk acknowledgement. 4. Press Create, then sign in and approve. Works on the web today. The plugin directory lives at chatgpt.com/plugins. ## Connect Stax.ai to Cursor Endpoint: https://api.elaichi.ai/mcp 1. Open `~/.cursor/mcp.json`. 2. Add the endpoint under `mcpServers`. 3. Reload Cursor, then sign in and approve. Set up per machine, so repeat it on each computer you work from. ## Connect Stax.ai to any MCP client Endpoint: https://api.elaichi.ai/mcp 1. Add the endpoint as a remote MCP server. 2. Sign in and approve. The Elaichi Agent already has these tools, with nothing to set up. ## What the consent screen decides Only Read is granted by default, which is not enough to call a Stax.ai tool. Over MCP there is no trusted place to confirm a write in the moment, so the consent screen is the standing approval rather than a formality. Grant Read and Run tools. Think hard before granting Delete, which reaches into connected apps and cannot be undone. ## What teams do with Stax.ai through Elaichi ### File the month's paperwork into stacks Accounting. Ask for the invoices and statements that came in this month and have them moved into the right client stack, with duplicates flagged before anything is filed twice. ### Set up a new TPA plan record Retirement plan administration. When a new plan client signs, create the TPA plan in Stax.ai and check the existing plan records so nothing is entered twice. ### Check what processing is still running Operations. See which document jobs are in progress and which finished, so you know when a batch is ready for review. ### Pull a document out of the archive Compliance. Search archived documents for a specific plan year or client, restore the one an auditor asked for, and download it for the response. ### Find what a client sent this week Client services. List recently added documents for a client, open the one they are asking about, and confirm which stack it landed in. ### Clean up stacks after a project closes Office administration. Archive the documents in a finished stack, copy anything that belongs elsewhere, and update the stack name so the workspace stays tidy. ## Frequently asked questions ### How do I connect Stax.ai to Claude? Connect Stax.ai in Elaichi by pasting your Stax.ai API key when asked, then in Claude open Customize, then Connectors, then Add, and paste the endpoint https://api.elaichi.ai/mcp. There is no OAuth application to register and no client ID or secret to generate. Once both steps are done, Claude can work with your Stax.ai stacks, documents, and TPA plans. ### Does Stax.ai work with ChatGPT and Cursor as well as Claude? Yes. Once Stax.ai is connected in Elaichi, the same endpoint, https://api.elaichi.ai/mcp, works in Claude, ChatGPT, Cursor, any MCP client, and the Elaichi Agent. You connect Stax.ai once and every client you use picks it up. ### What can an AI agent actually do with my Stax.ai data? An agent connected to Stax.ai can list and open your stacks and documents, upload new files, move or copy documents between stacks, archive and restore them, download a file you need, and find duplicates or recently added paperwork. It can also look up, create, and update TPA plans and check on document jobs still processing. Short, concrete asks such as "find last month's statements for Acme" work better than long paragraphs. ### Does connecting Stax.ai give the AI access to every stack and document? No. Access follows the person who signed in, so the AI only sees the Stax.ai stacks, documents, and TPA plans that person's own Stax.ai access allows. Elaichi can narrow that further with restrictions, but it can never widen it beyond what the person already has. ### Can my team share one Stax.ai connection? Yes. One person connects Stax.ai in Elaichi and shares the connection with a team, and nobody else ever handles the API key. Each teammate still signs in to Elaichi as themselves, so the audit log names the actual person who moved a document or created a TPA plan. ### Can I stop an agent from deleting or changing things in Stax.ai? Yes. In Elaichi you can block individual actions for Stax.ai, such as deleting documents, deleting stacks, or updating TPA plans, while leaving reading and filing allowed. A blocked action is never shown to Claude, ChatGPT, Cursor, or any other client, so no prompt can reach it. ### What happens to a Stax.ai connection when someone leaves? Offboarding a person in Elaichi ends their access to Stax.ai through every client at once. A shared Stax.ai connection keeps working for everyone else on the team. If you want to remove Stax.ai entirely, disconnecting it once in Elaichi removes it from Claude, ChatGPT, Cursor, and every other client. ## All 32 Stax.ai tools Every tool below is callable through https://api.elaichi.ai/mcp once Stax.ai is connected, subject to the toolbox it is in and the restrictions on the caller. - **List all Stax.ai stacks** (List). List all stacks for the current team in stax. Returns: id. - **Get single Stax.ai stack by ID** (Get). Get a single stax stack by identifier. Returns: id. - **Create a Stax.ai stack** (Create). Create a new stax stack for the current team. Returns: id. - **Update a Stax.ai stack by ID** (Update). Update an existing stax stack. Returns: id. - **Delete a Stax.ai stack by ID** (Delete). Delete a stax stack by identifier. Returns an empty 204 response on success. - **List all Stax.ai stack documents** (List). List documents in a stax stack. Returns document records from the stack with pagination and filtering support. Returns: id. - **List all Stax.ai documents** (List). Search documents in Stax using Elasticsearch. Returns document objects including id and content-type-specific attributes; full field breakdown depends on the document's schema in Stax. - **Get single Stax.ai document by ID** (Get). Get a single document in Stax. Returns the document object including id and content-type-specific attributes; full field breakdown depends on the document's schema in Stax. - **Create a Stax.ai document** (Create). Create a new document in Stax. Returns the created document object including id and content-type-specific attributes; full field breakdown depends on the document's schema in Stax. - **Update a Stax.ai document by ID** (Update). Update an existing document in Stax. Returns the updated document object including id and content-type-specific attributes; full field breakdown depends on the document's schema in Stax. - **Delete a Stax.ai document by ID** (Delete). Delete a document in Stax.ai. Requires docId. Returns success field indicating if the document was removed successfully. - **Stax.ai documents upload** (Upload). Upload one or more documents (up to 10 files) to Stax.ai. Returns an uploads array with key document fields such as _id, name, format, download URL, and metadata. - **Stax.ai documents move** (Move). Move a specific document in Stax.ai to another stack. Requires docId and stackId. Returns a success flag indicating whether the operation was successful. - **Stax.ai documents copy** (Copy). Copy a document to another stack in Stax.ai. Requires docId and stackId in the request body. Returns document details including _id, name, stack, format, and metadata fields. - **Stax.ai documents archive** (Archive). Archive a document in Stax.ai using the required docId. Returns the success field indicating whether the document was successfully archived. - **Stax.ai documents restore** (Restore). Restore an archived document in Stax.ai. Requires docId. Returns success field indicating whether the restore operation was completed successfully. - **Stax.ai documents download** (Download). Download the original document file from Stax.ai using id. Returns the document file in binary format as application/pdf or application/octet-stream. - **List all Stax.ai recent documents** (List). List recently viewed documents in Stax. Returns: id. - **List all Stax.ai document jobs** (List). List document_jobs in Stax.ai. Returns job details including _id, team, user, success, working, error, worker, queueTime, startTime, endTime, jobType, jobId, and jobData for each job. - **List all Stax.ai document duplicates** (List). Find duplicate documents in stax by comparing a document's checksum against all stored documents. Returns an array of matching Document objects; each item includes id. Required: docId. - **List all Stax.ai archive search** (List). Search archived documents in stax. Returns a list of archived document records from the results collection; the exact field-level shape of each record is document-specific and not enumerated in the available source — consult the stax API docs for the full record schema. - **List all Stax.ai tpa plans** (List). List all TPA plans in stax. Returns: planId, externalReference. - **Get single Stax.ai tpa plan by ID** (Get). Get TPA plan in Stax.ai by id. Returns planName, planType, active status, internalPlanId, pensionProId, directory, notes, and lastModified fields for the specified plan. - **Create a Stax.ai tpa plan** (Create). Create a new TPA plan in stax. Returns: planId, externalReference. - **List all Stax.ai tpa accounts** (List). List all TPA accounts for a plan in stax. Returns: accountId, accountNumber, planId, externalReference. Required: planId. - **Get single Stax.ai tpa account by ID** (Get). Get a single TPA account in stax by its account ID. Returns: accountId, accountNumber, planId, externalReference. - **Create a Stax.ai tpa account** (Create). Create a new TPA account in stax and associate it with a plan. Returns: accountId, accountNumber, planId, externalReference. Required: planId, accountNumber. - **List all Stax.ai tpa statements** (List). List all TPA statements in stax. Returns: statementId, planId, accountId, accountNumber, externalReference. - **Create a Stax.ai tpa statement** (Create). Attach a new TPA statement in stax. Returns: statementId, planId, accountId, accountNumber, externalReference. - **Stax.ai tpa statements confirm** (Action). Confirms (locks) a TPA statement by its ID, preventing further modifications to the statement. The endpoint accepts a POST request to /tpa/statement/confirm with a statement ID and requires cookie or bearer token authentication. - **List all Stax.ai tpa transactions** (List). List TpaTransaction records for a statement in stax. Returns: id, attributes (transaction-specific fields defined in the stax Trust Accounting API schema; each TpaStatement has many TpaTransaction records). - **Create a Stax.ai tpa transaction** (Create). Create a new TpaTransaction record in stax. Returns the created transaction including id and attributes (transaction-specific fields defined in the stax Trust Accounting API schema).