MindStudio vs StackAI (2026)

Stands at F on an enumerated catalogue rather than a headline count, which makes this one of the few Int cells in the index that can be checked rather than believed. Breadth across classes is the axis test and it is met decisively: mail and calendar, chat across four platforms, CRM and enrichment, documents, databases, storage, social, payments and media all have named blocks. The deepest single ecosystem is Google at 35 blocks, but the catalogue is not Google-shaped overall, which is what separates this from the single-ecosystem depth that takes a vendor to P. Two escape hatches keep the ceiling open: HTTP Request for any endpoint and a connector registry for raw third-party APIs. Recorded under section 7: the blocks that call Zapier, Make and n8n are MindStudio reaching out to those platforms, which counts as its integration breadth here, and is not evidence of MindStudio's own extensibility, graded separately on Ext.

Stands at F on an enumerated catalogue of roughly ninety individually documented app nodes, each with its own actions, inputs and outputs, which makes this checkable rather than a headline number. Breadth across classes is emphatic and skews enterprise in a way that distinguishes it from the prosumer catalogues elsewhere in this lane: SAP, NetSuite, Oracle, Workday, ServiceNow, Snowflake, Databricks and Egnyte are not integrations a no-code tool aimed at individuals carries. Eight database and warehouse connectors with natural-language-to-SQL querying is a second distinguishing cluster. The MCP node means anything exposed as an MCP server becomes callable too, so the ceiling is open. Recorded per section 7: the Zapier and Make nodes are StackAI reaching out to those platforms, which counts as its own integration breadth here and is not evidence of StackAI's extensibility, graded separately on Ext.

Stands at F. Both halves of the axis are documented as blocks rather than claimed in prose. Multi-step control flow is unusually complete for a no-code builder: AI-evaluated routing, comparison-operator branching, menus, jumps between workflows preserving scope, and escape hatches into JavaScript, Python or inline TypeScript when the visual surface runs out. Multi-agent is genuine and bidirectional: a workflow spawns one or more CHILD workflows, and separately a workflow published as a packaged workflow is callable by another MindStudio agent, so composition works in both directions with explicit input and output schemas. Snippets of reusable node groups and packaged custom blocks let a team build its own vocabulary on top, which is closer to organisational reuse than to templating.

Stands at F. Both halves are documented. Multi-step control flow is complete: AI Routing for model-decided branching, If/Else for deterministic branching, Loop Subflow for iteration, Delay for pacing, and Code and Python nodes as escape hatches. Multi-agent is genuine and works two ways, which is the part worth carrying: Subflow Tools let an AI Agent node call another flow AS A TOOL, so a supervising agent selects among specialist subflows at runtime, and the StackAI Project Node lets one project invoke another as a unit. A dedicated Orchestrating AI Agents guide sits in the tips tree, and Handling Errors and Fallback plus Skip and Replace Node give the reliability affordances that separate a production orchestrator from a demo canvas.

Stands at F and is now one of the best-evidenced Trig cells in the index, because the vendor documents the entire trigger surface in one place as an enumerated set of run modes rather than scattering it across product pages. All three classes the axis asks for are present with configuration detail: events through inbound email and webhook, schedules through CRON with timezone, and channels through the Chrome extension, the on-demand interface and agent-to-agent invocation. Two details worth carrying. The email mode restricts invocation to an approved sender list, which is a trigger-level access control. And packagedWorkflow mode makes an agent callable by an MCP service, meaning an external assistant can invoke a MindStudio agent as a tool, which is credited to Ext rather than double-counted here.

P>F. Channel coverage is the strongest part and is documented page by page rather than claimed: eight distinct end-user surfaces, which is wider than anything else reviewed in this lane. Events are covered by a dedicated Trigger node plus app-level triggers such as inbound email and inbound webhooks. Schedules are the thinnest of the three classes and the reason confidence is medium rather than high: recurring execution is referenced in two separate documentation pages as an established pattern, but no dedicated scheduling page appears in the complete index, so the configuration surface was not read. Fetching the Trigger Node page would settle it and is the single call that would take this to high confidence. Graded F because all three classes are documented as shipped rather than because the schedule surface is fully described.

P>F. The April P understated this substantially. Retrieval here is not a single ingest-and-query feature but a full managed lifecycle exposed as workflow blocks, so a builder can create, populate, inspect and tear down knowledge bases programmatically inside an agent rather than only through a settings screen. Relevant to the open Knowledge convention: this is decisively a maintained retrieval structure, not per-request assembly, since data sources are durable named artifacts with their own CRUD surface. The distinguishing detail worth carrying to comparison pages is that the customer also chooses the EMBEDDING and RERANKING models from the same catalogue that serves the generation models, which almost nothing else in this lane exposes, and which means retrieval quality is tunable rather than fixed by the vendor.

Stands at F. Knowledge bases are durable managed artifacts with their own creation, usage, node and feature documentation, plus a REST endpoint and, importantly, their own PERMISSION model, which is rare: access to a knowledge base is restricted independently of access to the workflow that uses it. For a regulated buyer that is the control that makes a shared agent platform viable across departments. Relevant to the open Knowledge convention: this is decisively a maintained retrieval structure rather than per-request assembly, and the vendor documents two distinct ways an agent consumes it. A Dynamic Vector Store covers the runtime-constructed case, and Search Connected Apps grounds on live systems without indexing them first, so both patterns are available and separable. Customer-managed vector stores through Pinecone and Weaviate mean the index can live outside the platform entirely.

RESOLVED FROM U to P, and deliberately not higher. State genuinely persists between runs, which rules out N: the Check for Changes block compares the current input against previously stored state, so the platform retains something across invocations by design, and an app-managed SQL database gives builders a durable store they control. What is absent is memory as the axis's F bar defines it. There is no memory module, no conversational memory that carries across sessions on its own, and nothing that learns from prior runs. This is the same shape that took charm, zed, blitzy and moderne to P in Coding agent, persisted state without accumulation, and it sits clearly below the F earned by cosine, greptile and cognition for agents that improve by reading their own history. Graded from a complete enumerated catalogue of 200 blocks rather than from a failed search, which is what makes P safe rather than provisional.

Stands at P, and it is now the ONLY cell below F on this record, so the reasoning matters. The Shared Memory node is explicitly a windowed context-passing mechanism, not a memory store: the builder chooses how many past interactions to forward and the documentation warns that passing too many will overwhelm the receiving model's context window. That is the same token-buffer shape as dify, and it is bounded by the context window rather than by a persistence layer. Nothing in the complete documentation index describes memory that survives a session, accumulates over time, or is learned from prior runs, which is the F bar met by cosine, greptile and cognition. Two things were deliberately NOT counted here, to avoid one fact doing work on three axes: the browser sandbox persists cookies and session state across runs, which is graded on Comp, and Canvas keeps per-conversation version history, which is a document feature. The Dynamic Vector Store is a retrieval structure and is graded on Know.

RESOLVED FROM U to F. The Checkpoint block is the decisive evidence and it is unambiguous: it pauses the run, hands the workflow state to a person, and lets them REVISE it before execution continues, which is more than the approve-or-reject pattern most vendors ship. It is the vendor's own mechanism inside the vendor's own runtime, so it clears the 30 August ruling about gates the customer already owns. The guardrail half is separately strong and unusually concrete for a no-code platform: PII detection and redaction are shipped blocks implemented on Microsoft Presidio rather than a policy statement, and the payment blocks enforce a spend cap by design. Approved-sender lists on the email trigger are a fourth control, restricting who can invoke an agent at all.

Stands at F. Human in the Loop is a documented first-class node rather than a posture, and the vendor treats it as foundational enough to build the first of its five learning challenges around it, which is a good signal that it is a load-bearing feature rather than a checkbox. The approval surface is the vendor's own and is delivered where the approver already is: the product page shows an approval prompt in Slack with explicit approve and disapprove controls. That clears the 30 August ruling, since Slack is the delivery channel while the gate itself belongs to StackAI. Workflow Notifications and the production workflow lock in Project Controls are separable second and third oversight mechanisms, the latter unusual because it constrains what a builder can change rather than what an agent can do.

P>F, resolving the 31 August flag that this was the most likely understated cell on the record. The Vanta trust centre still renders client-side, but the attestation is stated on a page that does read. THE ATTESTATION IS FIRST PARTY AND UNHEDGED: THE PLATFORM IS SOC 2 TYPE I AND II CERTIFIED. Certified, with both types named. That is above the hedge rungs that held altilia, aigensei and ren3 at Partial in this lane, and it removes the reason the previous basis gave for Partial, which was explicitly that NO FIRST-PARTY ATTESTATION STATEMENT WAS RETRIEVED. ONE CARE POINT, BECAUSE THIS PAGE IS PARTLY BUYER GUIDANCE. The same article contains generic advice about why enterprises demand SOC 2 and which frameworks apply, which is the shape refused on EverWorker, latenode and flowx-ai in this lane. The difference is that this page also makes self-descriptive claims in the vendor's own voice, MINDSTUDIO PROVIDES COMPREHENSIVE SECURITY FEATURES and THE PLATFORM IS SOC 2 TYPE I AND II CERTIFIED, and only those sentences are graded. The generic passages are not. THE CONTROL HALF IS DOCUMENTED SEVERAL TIMES OVER. Comprehensive audit logging tracking WHO CREATED WHICH AGENTS, WHEN THEY WERE DEPLOYED, WHAT DATA THEY ACCESSED, AND WHAT ACTIONS THEY PERFORMED, which is agent-level attribution rather than platform logging. GDPR data handling with consent management and deletion on request. Agent-level guardrails, content filtering and rate limiting to prevent runaway processes. THE SHIPPED BLOCKS FROM THE PREVIOUS BASIS REMAIN AND ARE STILL DISTINCTIVE: Detect PII and Redact PII using Microsoft Presidio, and Check App Role branching on whether the current user holds a given application role. A DPA, privacy policy and terms are published at first-party legal URLs. CONFIDENCE STAYS MEDIUM RATHER THAN HIGH FOR ONE REASON: trust.mindstudio.ai is a Vanta trust centre that renders client-side and has now failed retrieval on two separate passes, so the report itself, its period and its auditor were not read. The certification claim is the vendor's own statement rather than a readable artefact. SSO and SCIM are reported by third parties and are not credited here.

Stands at F and is among the two or three strongest Sec cells in the index. The conjunction is met several times over, and unusually the control surface is documented page by page rather than asserted as a bullet list: RBAC, workspace and folder isolation, feature-level admin enforcement, per-connection and per-knowledge-base permissions, MFA, and full SCIM provisioning with separate Okta and Entra guides. Encrypted environment variables resolved per stage is a maturity signal most no-code platforms lack. Graded strictly on the control surface: on-premise deployment and government cloud are carried on Dep and deliberately excluded here, per the 30 August ruling that sovereign delivery never reaches Security. Disclosure quality is high by index standards, with a trust centre, a SOC 2 report request route, model-provider DPAs published as signed PDFs and a BAA route, though no audit firm is named.

Stands at P. Run history exists and is user-visible, which rules out N, and the platform gives builders a way to label runs so a history is navigable rather than opaque. But nothing retrieved describes a trace: no per-block execution view, no variable inspection after the fact, no audit log, no export. Under the axis rule this is the reporting side of the line, seeing that something ran and what it was called, rather than reconstructing why it took the path it did. Notably, the platform HAS the raw material for a strong Obs story, since every workflow is a graph of discrete blocks, so a per-block trace would be natural to build and may well exist in-product without being documented on any surface reached. Confidence is medium for that reason. The Test and Evaluate documentation section is the likely home of it and is the same fetch that would settle Eval.

Stands at F but at MEDIUM confidence, and the gap is named rather than glossed. A dedicated Observability section with three pages, a REST Analytics endpoint, audit logs on the product page and version history through the development lifecycle is comfortably more than reporting, and the Evaluator adds a quality dimension most vendors have nothing equivalent to. What was NOT confirmed is the per-run trace: the Manager page was identified from the complete documentation index but not read, so whether a customer can reconstruct why a specific run took the path it did is inferred rather than verified. One detail cuts slightly against F and is recorded honestly: a guide titled Adding Advanced Logging for Analytics suggests some richer telemetry is assembled by the builder rather than supplied. Fetching the Manager and Analytics pages would settle this to high confidence either way, and it is the main outstanding item on this record.

RESOLVED FROM UNRESOLVED, P>F, and the record's internal contradiction is settled in favour of the longDescription. THE CONTRADICTION, RESTATED: longDescription claimed self-hosting while deploymentOptions carried SaaS only, and the April basis was an internal artifact with no page behind it. Two first-party sources now settle it and both say self-hosting. THE VENDOR ANNOUNCEMENT IS THE PRIMARY EVIDENCE. MindStudio for Enterprise was announced as an offering that BRINGS ALL THE POWER OF MINDSTUDIO'S HOSTED SAAS PLATFORM DIRECTLY TO CUSTOMERS' ON-PREMISE SERVERS OR PRIVATE CLOUDS, CREATING FULLY ISOLATED AND SECURE ENVIRONMENTS, with the vendor's own framing that YOUR EMPLOYEES' USE OF AI NEVER LEAVES YOUR NETWORK. Under the ground rules a vendor announcement clears the first-party floor and there is no recency floor, so the 2024 date goes to the note rather than the grade. THE CURRENT PRODUCT PAGE CONFIRMS IT IS STILL OFFERED: FOR ORGANIZATIONS WITH STRICT DATA RESIDENCY REQUIREMENTS, MINDSTUDIO OFFERS SELF-HOSTED DEPLOYMENT OPTIONS. THIS ALLOWS ENTERPRISES TO RUN THE ENTIRE PLATFORM WITHIN THEIR OWN INFRASTRUCTURE, ENSURING DATA NEVER LEAVES THEIR CONTROL. Entire platform, not an agent or a connector, which is the distinction that kept make at Partial on its on-prem agent earlier today. PRIVATE MODELS ARE THE PART THAT CLOSES THE LOOP and few records in this lane manage it. The enterprise offering carries ADDITIONAL SUPPORT FOR PRIVATE MODELS, so inference runs inside the deployment rather than being sent to a hosted provider. A self-hosting claim whose model calls still leave the perimeter answers only half the question; this one does not have that hole. COMMERCIAL PLACEMENT: self-hosting sits on the Business plan alongside custom domains and model access control. What is not documented on any first-party page reached is region selection within the hosted offering, or installation requirements and supported infrastructure for the self-hosted path. Customer-controlled deployment is the stronger half of this axis and it is met; the hosted residency half is not.

Stands at F, and the strongest single piece of evidence is not a claim but an artifact: StackAI ships a dedicated enterprise CLI whose entire purpose is deploying and managing the platform on customer infrastructure, with its own documentation set, engineering standards, Kubernetes and Docker migration guides and a CVE upgrade runbook. A vendor that maintains a deployment CLI with release engineering docs is not offering on-premise as a sales concession. Government deployment is a separate documented path including Azure OpenAI in Azure Government, which is a genuine sovereignty story rather than legal cover. Per the 30 August ruling all of this lives here and only here; the Security cell was written on the control surface alone. Local LLM hosting reinforces it, since an air-gapped-leaning customer can keep both the platform and the inference inside its own boundary.

Stands at F, and the basis is now specific where the April one named nothing. The unusual thing here is WHERE the prebuilt work lives: rather than a separate template gallery, roughly twenty vendor-authored packaged workflows sit inside the block catalogue itself, so a builder drops a finished capability such as Deep Research or Analyze CSV straight onto the canvas as a single step. That is a stronger form of prebuilt than a template you copy and adapt, because the packaged workflow stays maintained by the vendor. Clears the browsable-catalogue bar carried from Coding agent by a different route: the catalogue is the product surface, publicly browsable and enumerated. Customers publish their own packaged workflows as custom blocks alongside them.

Stands at F and the basis is now specific where the April one named nothing. Three separable layers of prebuilt material, which is more than most: a browsable Templates catalogue reachable from both the product and the documentation, named agent templates such as Content Writer that the documentation directs users to open and adapt, and Skills, which are reusable instruction packs agents load on demand and which function as prebuilt behaviour rather than prebuilt structure. A Prompt Library and a Common Architectures page add reusable material at the prompt and pattern level. Clears the browsable-catalogue bar carried from Coding agent. Recorded honestly: the catalogue page itself was not enumerated this pass, so its depth is unmeasured, though the vendor names specific templates by name in the documentation which is stronger than a bare claim of a library.

Stands at F, and this is the strongest Model cell reviewed anywhere in the index so far. It is not a claim about flexibility, it is an enumerated catalogue: 398 individually selectable models from 49 named publishers, each with its own page and published per-token or per-unit rate, spanning text, vision, image, video, 3D, transcription, speech, music, lip sync, embedding, reranking and document extraction. Selection is per block rather than per account, so a single workflow can reason on a frontier model and summarise on a cheap one. THE RECORD'S 200+ FIGURE IS STALE and has been corrected to 398+. The commercial mechanism matters as much as the count: no separate provider accounts or API keys, billed at the providers' own rates. Deprecated models remain listed and labelled, which is unusually honest catalogue hygiene and lets a builder see what is being retired.

Stands at F, and the distinguishing feature is not the breadth of model choice but the GOVERNANCE over it, which nothing else reviewed in this lane ships. LLM Provider Governance lets an administrator control which models the organisation may use and where information is sent and stored, so model flexibility is bounded by policy rather than left to whoever builds the workflow. For a regulated buyer that is the difference between model choice being an asset and being a compliance risk. Underneath it the ordinary requirements are met: multiple providers, per-node selection, models hosted in the customer's own Azure or Bedrock account, and locally hosted models. Not to be confused with the Ext credit for publishing an MCP server, which is the opposite direction of travel and is graded there.

Stands at F and clears Mike's 30 August Ext bar on both halves: a documented API for calling agents from outside, and a published SDK, which is exactly what zencoder and baz lacked when they were corrected down to P. The serverless framing is the useful detail, since an agent becomes an addressable function rather than something a customer has to host. MCP is present as corroboration rather than as the bar, and it runs in the credited direction: packagedWorkflow mode lets an external MCP service invoke a MindStudio agent, so the customer picks the assistant that reads the vendor, which section 7 says credits Ext and refuses Model. Inbound webhooks with secret protection and outbound custom code in three languages complete the surface.

Stands at F and clears Mike's 30 August Ext bar on every available route rather than just one. A documented REST API with named endpoints makes the platform callable from outside, which is the bar itself. StackAI also PUBLISHES ITS OWN MCP SERVER, which under the section 7 axis rule credits Ext and refuses Model, because the customer picks the assistant that reads StackAI. The enterprise CLI is a third surface, and project export and import means an agent definition is a portable artifact rather than something locked in a tenant. The Custom API node is the inbound counterpart, letting a workflow call anything the catalogue does not cover. This is a wider extensibility surface than most no-code platforms in this lane offer, and it is the reason the platform can sit underneath another product rather than only in front of a user.

F>P, the one downward correction on this record, and it is a levelling call rather than a finding of absence. Real testing exists: a dedicated workbench run mode for testing and development, a documented Test and Evaluate section of the University covering how to evaluate performance and debug issues before deployment, run history, and Set Variable for injecting test values in workbench mode. What is missing is the F bar as this lane has now set it. Lyzr earns F on automated test-case generation, domain scenario suites, production-readiness scoring and regression re-runs on model change; goose and openhands earned F in Coding agent on customer-runnable harnesses. Nothing of that kind was reached here: no scoring, no test suites, no regression, no benchmarking. The April F was graded from an uncheckable internal report and is not supported by anything retrieved. Fetching the Test and Evaluate section itself is the one call that would settle whether this should return to F.

P>F. The Evaluator is a shipped customer-facing evaluation product, not the vendor testing its own work, which is the distinction the axis turns on. LLM-as-a-judge scoring of the customer's own agents clears the bar set by goose and openhands in Coding agent and matches the shape that took lyzr to F earlier in this lane. The surrounding surface is stronger than the Evaluator alone: an Agentic Development Lifecycle with version control, pull requests and prompt diffs, Project Controls that track versions during development and LOCK a workflow once in production, and documented troubleshooting and error-fallback guides. The production lock is the detail worth carrying, since it is a governance control on change rather than a testing feature and few vendors in this lane ship one. Recorded honestly: the Evaluator page itself was identified from the complete documentation index and its description rather than fetched in full, so the scoring mechanism's depth is not measured.

Stands at N, and the reasoning is the axis error corrected thirteen times in Coding agent, applied here in advance. THREE THINGS ON THIS PLATFORM LOOK LIKE COMPUTER USE AND NONE OF THEM ARE. Screenshot URL captures a page as an image over HTTP; it renders, it does not operate. The LinkedIn and X scraping blocks read public pages programmatically. And the Chrome extension run mode, which is the most tempting, is a TRIGGER: the extension hands the workflow the current page's URL, metadata, text and the user's selection as launch variables, so the agent reads what a human is looking at rather than driving the browser itself. The axis is non-zero only when the agent operates software with no programmatic interface, and every external action here goes through an API, an HTTP request or a scrape. Confidence medium rather than high because the extension is close enough to the line that a later pass should confirm it cannot act on the page.

N>F, a full point, and the most consequential single correction of this lane so far. THE REASONING IS THE THIRTEEN JUNE COMP ERRORS RUN IN REVERSE, so this record is worth keeping as the worked example of the distinction. The StackAI Computer provider ships three actions and only one of them is Comp. THE TERMINAL TOOL IS EXPLICITLY NOT CREDITED HERE: running shell commands in an isolated sandbox is code execution, which is exactly what blink-new, codebuff, compyle, cosine, cubic and eight others were wrongly graded F for. Canvas is a document workspace and is likewise refused. What earns F is browser navigation, and it earns it decisively: the agent drives a real browser through authenticated web applications whose session state persists between runs, which is the definition of operating software with no programmatic interface. Two modes exist, a deterministic replay of a recorded sequence and an agentic mode for unfamiliar or dynamic tasks, and a separate HyperBrowser node adds a third route. A live stream URL for watching execution and a step-by-step result payload are unusual and make the capability inspectable rather than opaque.