MindStudio

The capability catalogue enumerates 200 blocks across seventeen categories: Google Suite 35 covering Docs, Sheets, Drive, Calendar, Gmail and search, Utilities 26, Video Editing 18, Workflow Logic 15, CRM and Enrichment 13 covering HubSpot, Apollo, Hunter.io and ActiveCampaign, Social Media 12 across LinkedIn, X and YouTube, Text Processing 11, Productivity Apps 11 covering Notion, Coda and Airtable, AI Generation 9, Databases 9, Telegram 8, Messaging 7 covering Slack, email, SMS and Discord, Image Editing 7, Web Data 6, AI Analysis 6, Security and Payments 4, and Automation Platforms 3. An HTTP Request block calls any external endpoint and a MindStudio Connector Registry block runs raw API connectors to third-party services.

Sourcemindstudio.ai capability catalogueread 2026-08-31

Workflow Logic ships Router for routing execution to different branches based on AI evaluation, comparison operators or workflow jumps, Run Workflow for spawning one or more child workflows, Jump for transitioning to another workflow while maintaining scope, Menu for branching on user choice, and Run Function, Run Inline Script and Run Packaged Workflow for custom JavaScript, Python or TypeScript and reusable custom blocks. The Start block's packagedWorkflow run mode makes an agent invocable by another MindStudio agent with defined input and output schemas, and the End block optionally returns structured output to the caller.

Sourcemindstudio.ai capability catalogue, Workflow Logic category and Start blockread 2026-08-31

The Databases category ships a complete vector lifecycle as workflow blocks: Query Data Source searches a vector store and returns relevant chunks, with Create, List and Delete Data Source, Upload, Fetch and Delete Data Source Document alongside it, plus Query External SQL Database and Query App Database.

Ingestion is served by Extract Text from URL with OCR for scanned PDFs, Scrape URL for structured web content, and document-extraction models including Mistral OCR, LlamaParse and Google Document AI. The model catalogue exposes selectable embedding models from Voyage AI, Cohere, OpenAI, Google and Qwen and dedicated reranking models from Cohere, Voyage and Qwen.

Sourcemindstudio.ai capability catalogue and model catalogueread 2026-08-31

Checkpoint is a first-class Workflow Logic block documented as allowing the user to review, and optionally revise, the state of the workflow before proceeding.

Guardrails ship as separate blocks: Detect PII and Redact PII scan and mask personally identifiable information using Microsoft Presidio, Check App Role branches execution on whether the current user holds a specific role, and Authorize Payment creates an authorization under a total cap that bounds all subsequent billing.

Menu and User Input provide additional human decision points mid-run, and the email run mode restricts invocation to an approved sender list.

Sourcemindstudio.ai capability catalogue, Workflow Logic and Security & Payments categoriesread 2026-08-31

The vendor's enterprise security page states in its own voice that MindStudio provides comprehensive security features for enterprise agent deployment and that the platform is SOC 2 Type I and II certified. Named customer-facing controls include comprehensive audit logging tracking who created which agents, when they were deployed, what data they accessed and what actions they performed; GDPR data handling with consent management and deletion on request; and agent-level guardrails, content filtering and rate limiting.

Shipped security blocks include Detect PII and Redact PII using Microsoft Presidio and Check App Role branching on application role. A Data Processing Addendum, privacy policy and terms of use are published at first-party legal URLs. The same page also contains generic buyer guidance about SOC 2 and compliance frameworks, which is not graded. The Vanta-hosted trust centre at trust.mindstudio.ai renders client-side and has failed retrieval on two passes, so the report period and auditor were not read; SSO and SCIM are reported only by third parties and are not credited.

Sourcemindstudio.ai enterprise security page and capability catalogue, first-party legal pagesread 2026-09-01

Set Run Title is a documented block that sets the title of an agent run for the user's history, which establishes that runs are recorded and browsable. Workspace management documentation covers usage controls and billing settings, and a Test and Evaluate documentation section references debugging issues. No execution trace, per-block run inspection, audit log, or log export appears in the complete 200-block capability catalogue or in the documentation structure reached, and the vendor's error handling surfaces no post-run diagnostic view.

Sourcemindstudio.ai capability catalogue and university.mindstudio.ai overviewread 2026-08-31

Check for Changes detects differences between runs by comparing current input against previously stored state and routes execution on whether a change occurred, which is documented cross-run state retention. Set Variable and Build User Context persist values within a run, Query App Database executes SQL against an app-managed database, and Set Run Title labels a run for the user's history. No memory module, cross-session conversational memory or learned-memory feature appears anywhere in the complete 200-block capability catalogue across its seventeen categories.

Sourcemindstudio.ai capability catalogueread 2026-08-31

The vendor's enterprise security page states that for organizations with strict data residency requirements MindStudio offers self-hosted deployment options allowing enterprises to run the entire platform within their own infrastructure so data never leaves their control.

The vendor's announcement of MindStudio for Enterprise describes bringing the full hosted SaaS platform to customers' on-premise servers or private clouds as fully isolated environments, with additional support for Private Models so inference runs inside the deployment. Self-hosting is placed on the Business plan alongside custom domains and model access control.

No region selection within the hosted offering, installation requirement or supported infrastructure specification is documented on any first-party page reached.

Sourcemindstudio.ai enterprise security page, MindStudio for Enterprise announcementread 2026-09-01

The capability catalogue contains roughly twenty vendor-authored packaged workflows exposed as single blocks, including Deep Research, Analyze CSV, OCR Image, Create Podcast, Track Website Changes, Extract from URL, Generate Spritesheet, and a Text Processing set of Summarize Text, Improve Writing, Translate Text, Rewrite for Audience, Convert Reading Level, Merge Text, Adjust Tone, Align to Brand, Rewrite Text, Convert to Format and Extract from Text. A Run Packaged Workflow block lets customers publish and reuse their own custom blocks on the same footing, and the Start block's packagedWorkflow mode defines the metadata and schemas that make one shareable.

Sourcemindstudio.ai capability catalogueread 2026-08-31

The Start block documents seven run modes determining how a workflow is invoked: onDemand for manual user launch, workbench for testing, schedule for CRON-based execution with a timezone and optional static launch variables, email where inbound mail to a dedicated address triggers the run and populates sender, subject and body, extension where a Chrome extension triggers from the currently open page supplying URL, metadata, page text and selected text, webhook where a POST to a dedicated secret-protected URL passes structured data, and packagedWorkflow for invocation by another MindStudio agent or an MCP service with defined input and output schemas. Launch variables carry data in at invocation time for all non-manual modes.

Sourcemindstudio.ai capability catalogue, Start blockread 2026-08-31

The model catalogue lists 398 models across 49 publishers, each individually selectable with a published rate, including Anthropic 13, OpenAI 49, Google 45, Qwen 34, X.ai 28, Mistral 18, ByteDance 18, Wan 15, Black Forest Labs 13, WaveSpeed 13, Perplexity 11, Kling 11, Meta 10, DeepSeek 9, Z.ai 9, Voyage AI 8, Ideogram 8, Luma Labs 7, Cohere 6, MiniMax 6 and others, filterable by modality across text, vision, image, video, 3D, transcription, text to speech, music, lip sync, embedding, reranking and document extraction. The vendor states no API keys are required.

Sourcemindstudio.ai model catalogueread 2026-08-31

The documentation carries an API Reference for integrating MindStudio agents into external applications, described as an API for serverless AI function calling, alongside a published NPM package for integrating agents into Node.js projects and a Custom Functions surface for running JavaScript or Python inside workflows. The End block optionally returns structured output for API callers, the Start block's webhook run mode accepts a secret-protected POST from any external system, packagedWorkflow mode exposes an agent to another agent or an MCP service with defined schemas, and Run Inline Script executes TypeScript directly in a workflow.

Sourceuniversity.mindstudio.ai developer section and mindstudio.ai capability catalogueread 2026-08-31

The Start block documents a workbench run mode explicitly for testing and development, in which Set Variable blocks supply test values in place of launch variables, and the University documentation carries a Test and Evaluate section described as covering tools and techniques for testing agents, evaluating performance, debugging issues and ensuring quality standards before deployment.

Set Run Title records runs into a user-visible history. No evaluation scoring, test-case generation, regression suite or benchmarking surface appears in the complete 200-block capability catalogue or in the documentation structure reached.

Sourceuniversity.mindstudio.ai overview and mindstudio.ai capability catalogueread 2026-08-31

No browser-driving, desktop-control or computer-use block appears in the complete 200-block capability catalogue across its seventeen categories.

The nearest capabilities are Screenshot URL, which captures a web page as a PNG, Scrape URL and the LinkedIn and X scraping blocks, which extract text and structured content over HTTP, and the Chrome extension run mode, which triggers a workflow from the currently open page and supplies its URL, metadata, full page text and any user-selected text as launch variables. All external actions are performed through connectors, HTTP requests or scraping, which are programmatic interfaces.

Sourcemindstudio.ai capability catalogue and Start blockread 2026-08-31