# Lobstr MCP connector The Lobstr connector lets Claude, ChatGPT, Cursor, or any MCP client start crawlers, queue tasks, check runs, and download results from your Lobstr account through Elaichi, with each person signed in as themselves. Source: https://elaichi.ai/connectors/lobstr/ ## Facts | | | | --- | --- | | Application | Lobstr | | Category | Marketing Automation | | AI tools | 42 | | 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. 42 tools is past the 30-tool threshold, so clients use `search_tools` and `execute_tool` | ## What you can ask once Lobstr is connected - Which Lobstr crawlers created the most tasks this week? - Create a Lobstr squid for the pricing page crawler. - Show tasks from every Lobstr squid paused since Monday. ## Connect Lobstr in Elaichi This happens once for the organization, before any client is involved. 1. Open Connections, choose Add connection, and pick Lobstr. 2. Optionally set Share with, then press Connect. 3. Paste a Lobstr API key. One person generates a token in Lobstr 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 Lobstr tools are exposed, rename them, or freeze arguments before anyone points a client at it. ## Connect Lobstr 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 Lobstr 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 Lobstr 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 Lobstr 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 Lobstr 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 Lobstr through Elaichi ### Queue a fresh lead list before Monday Sales. Ask for a new squid on the Google Maps or LinkedIn crawler, hand it the search URLs as tasks, and start the run so the results are waiting when the team logs in. ### Check whether last night's run finished Marketing. Ask how a run went, see how many results it collected and how many credits it used, and grab the download without opening Lobstr. ### Add new targets to an existing squid Growth. Paste a handful of URLs or search pages and have them added as tasks to the squid that is already set up, then kick off a run. ### Keep an eye on the credit balance Operations. Ask what the account balance is and which recent runs spent the most credits, so nobody discovers an empty balance the morning a campaign starts. ### Find the right crawler for a source Research. Ask which crawlers are available in Lobstr and what parameters each one takes before committing to a new squid. ### Tidy up finished squids and stale tasks Data. List the squids and tasks the team no longer needs and remove them, with every deletion recorded against the person who asked. ## Frequently asked questions ### How do I connect Lobstr to Claude? First connect Lobstr in Elaichi: you paste the API key from your Lobstr account settings, and that is the whole sign-in step, with no client ID or secret to generate. Then in Claude go to Customize, then Connectors, then Add, and paste the endpoint https://api.elaichi.ai/mcp. Claude signs you into Elaichi, and from then on it can work with your Lobstr crawlers, squids, tasks, and runs. ### Does Lobstr work with ChatGPT and Cursor as well as Claude? Yes. Once Lobstr is connected in Elaichi, the same endpoint, https://api.elaichi.ai/mcp, works in Claude, ChatGPT, Cursor, any other MCP client, and the Elaichi Agent. You connect Lobstr once and every client you use picks it up. ### What can an AI agent actually do with my Lobstr data? It can list the crawlers available in Lobstr and their parameters, create or update squids, add tasks such as URLs or search pages, start runs, and report on how a run went, including results collected, credits spent, and the download. It can also check your account balance and remove squids, tasks, or runs you no longer need. Because Lobstr exposes a long list of actions, short concrete asks like "start a run on the Paris restaurants squid" work better than long paragraphs. ### Does connecting Lobstr give the AI everything in my account? No. Every call runs inside the access of the person who signed in, so the agent only sees the Lobstr squids, runs, and balance that person can already see in Lobstr. Elaichi can narrow that further with restrictions, and it can never widen it beyond what the Lobstr account allows. ### Can my team share one Lobstr connection? Yes. One person connects Lobstr 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 when a run is started or a squid is deleted, the audit log names the person who did it. ### Can I stop an agent from deleting or changing things in Lobstr? Yes. Restrictions in Elaichi apply per action, so you can allow reading crawlers and runs while blocking the deletion of squids, tasks, or runs in Lobstr. A blocked action is never advertised to the AI client at all, so no prompt, however worded, can reach it. ### What happens to a Lobstr connection when someone leaves? Offboarding a person in Elaichi ends their access to Lobstr through every client at once. If they connected Lobstr and shared it with a team, the shared connection keeps working for everyone else. When you do want Lobstr gone, disconnecting it once in Elaichi removes it from Claude, ChatGPT, Cursor, and every other client together. ## All 42 Lobstr tools Every tool below is callable through https://api.elaichi.ai/mcp once Lobstr is connected, subject to the toolbox it is in and the restrictions on the caller. - **List all Lobstr plants** (List). List all plants in lobstr that the user has access to. Returns: id. - **Create a Lobstr plant** (Create). Create a new plant in the lobstr store. Returns: id. - **Delete a Lobstr plant by ID** (Delete). Delete a single plant in lobstr by id. Returns an empty 204 response on success. Required: id. - **List all Lobstr user** (List). Get the authenticated lobstr user's complete profile in a single call. Returns: personal_details, account_status, subscription. Useful as a first call after authentication to verify credentials and inspect account state. - **Get single Lobstr balance by ID** (Get). Get the current lobstr account balance and credit information. Returns: name, available, consumed, bonus_credit, rollover_credit, interval, used_slots, total_available_slots, has_unpaid_bill, reset_time. - **List all Lobstr crawlers** (List). List all available crawlers on the lobstr platform. Returns: id, name, is_available, is_premium, credits_per_row for each crawler. Use this to discover which crawlers are available before creating squids or adding tasks. - **Get single Lobstr crawler by ID** (Get). Get a single lobstr crawler by id. Returns: id, name, is_available, is_premium, credits_per_row. Required: id. - **Get single Lobstr crawler param by ID** (Get). Get the configurable input parameters for a specific lobstr crawler. Returns the full parameter configuration organized into two sections: task and squid. Required: crawler_hash. - **List all Lobstr squids** (List). List all lobstr squids belonging to the authenticated user. Returns squid objects including id and name. Optionally filter results by name using a case-insensitive partial match. - **Create a Lobstr squid** (Create). Create a new squid in lobstr. Returns the created squid object including id and name. - **Get single Lobstr squid by ID** (Get). Get a single lobstr squid by id. Returns: id, name. Required: id. - **Update a Lobstr squid by ID** (Update). Update an existing lobstr squid by id. Returns the updated squid object including id and name. Required: id. - **Delete a Lobstr squid by ID** (Delete). Delete a lobstr squid by id. Returns an empty 204 response on success. Required: id. - **List all Lobstr tasks** (List). List lobstr tasks. Returns: id, status, and params for each task record in the collection. - **Create a Lobstr task** (Create). Create lobstr tasks by adding URLs or search queries to a squid for scraping. Returns tasks (array with id, status, params) and duplicated_count indicating how many duplicate tasks were skipped. Required: squid, tasks. - **Get single Lobstr task by ID** (Get). Get a single lobstr task by id. Returns: id, status, and params containing the scraping parameters for the task. Required: id. - **Delete a Lobstr task by ID** (Delete). Delete a lobstr task by id. Returns an empty 204 response on success. Required: id. - **List all Lobstr runs** (List). List lobstr runs for a specific squid to view run history and past executions. Returns: id, status, is_done, total_results, credit_used, done_reason. Required: squid. - **Create a Lobstr run** (Create). Create a new run in lobstr. Returns: id, status, is_done, total_results, credit_used, done_reason. - **Get single Lobstr run by ID** (Get). Get a single lobstr run by id. Returns: id, status, is_done, total_results, credit_used, done_reason. - **Get single Lobstr run stat by ID** (Get). Get detailed real-time statistics for a specific lobstr run. Returns: percent_done, total_tasks_done, total_tasks, results_done, results_total, duration, eta, and is_done. Useful for polling run progress until the run completes. Required: run_hash. - **Get single Lobstr run credit by ID** (Get). Get credit usage details for a lobstr run by its hash identifier. Returns: run_id, total_credits, total_results, breakdown (per-function credit and attempt counts), rates, base_function_label, and filter_breakdown. Required: run_hash. - **Get single Lobstr run download by ID** (Get). Get a temporary download URL for a lobstr run's exported results file. Returns: s3 (a pre-signed temporary URL pointing directly to the data file). Required: run_hash. - **List all Lobstr run tasks** (List). List run tasks in lobstr. Returns the collection of task execution records. Returns: id, status, where status reflects the task lifecycle (pending, running, uploading, paused, done, aborted, error). - **List all Lobstr results** (List). List scraped data results from lobstr runs, queryable per squid. Returns an array of scraped data rows; the fields within each row are scraping-configuration-specific and vary by squid setup. To export results as a file, use the Download Run endpoint instead. - **List all Lobstr accounts** (List). List all connected platform accounts in lobstr. Returns a paginated array of account objects, each including: id, username, type, status, status_code_info, and baseurl. - **Get single Lobstr account by ID** (Get). Get a single lobstr connected platform account by id. Returns: id, username, type, status, status_code_info, and baseurl. Required: id. - **Delete a Lobstr account by ID** (Delete). Delete a lobstr connected platform account by id. Returns an empty 204 response on success. Required: id. - **Create a Lobstr account** (Create). Create a new connected platform account in lobstr. Returns: id, username, type, status, status_code_info, and baseurl. - **Create a Lobstr account cooky** (Create). Create a lobstr platform account by submitting authentication cookies for a specific account type. Returns the created account object including id, username, type, status, status_code_info, and cookies. Required: type, cookies. - **List all Lobstr account types** (List). List all available platform account types in lobstr and their authentication requirements. Returns: id, name, domain, baseurl, cookies (required credential names), and params (optional rate-limit and batch configuration). No required parameters. - **Get single Lobstr synchronize by ID** (Get). Check the synchronization status of a cookie update task in lobstr by its ID. Returns: id, object, status_code, status_text, account_hash. Required: id. - **Get single Lobstr crawler attribute by ID** (Get). Get the output attributes defined for a specific lobstr crawler. Returns the list of result fields the crawler can produce, including each attribute's name, type, description, function group, is_main flag, and example value. Required: crawler_hash. - **Get single Lobstr squid chain by ID** (Get). Get the chain configuration for a lobstr squid. Returns the downstream chain settings including target_module_id, field_map, autostart, cluster_name, and cluster_concurrency. Required: squid_hash. - **Create a Lobstr squid chain** (Create). Configure a chain on a lobstr squid so its results automatically feed into a target squid. Returns the chain configuration including target_module_id, field_map, autostart, cluster_name, and cluster_concurrency. Required: squid_hash, target_module_id. A squid can only have one outgoing chain at a time; chains are limited to a maximum depth of 3 squids. - **Delete a Lobstr squid chain by ID** (Delete). Delete the chain configuration on a lobstr squid, removing the downstream link to the target squid. Returns an empty 204 response on success. Required: squid_hash. - **Create a Lobstr squid empty** (Create). Empty a lobstr squid by removing all its tasks and results while preserving its configuration. Returns: id, object, emptied, deleted_count. Required: squid_hash. - **Create a Lobstr task upload** (Create). Upload tasks to lobstr to queue them for scraping. Returns: tasks (array of task objects each with id, status, and params) and duplicated_count indicating how many submitted tasks were duplicates. - **Get single Lobstr task upload by ID** (Get). Get a single lobstr task upload by id. Returns: id, status, and params containing the target URL. Required: id. - **Create a Lobstr run abort** (Create). Abort an active lobstr run before completion using its hash ID. The run stops collecting data immediately and begins exporting any results already gathered. Returns: id, status, is_done, done_reason, total_results, duration, credit_used. Required: run_hash. - **Create a Lobstr delivery** (Create). Create and configure a delivery integration for a lobstr squid (scraping job). Supports delivery methods including email, Google Sheets, webhook, S3, and SFTP. The source documentation for this endpoint does not enumerate specific parameter names or response fields. - **Create a Lobstr delivery test** (Create). Test a lobstr delivery configuration for a specified delivery method. Returns an empty 204 response on success. Required: route (one of 'email', 'googlesheet', 'webhook', 's3', or 'sftp').