SmartStaff.io vs Microsoft Copilot Studio

Can a person read that plan before the work starts?

The activity map 'shows a visual representation of the plan that the agent generated' — live or retrospective trace. Its Edit button edits the component definition, not the run's plan.

Can a person rewrite, reorder or drop steps and then approve?

Out of the box, does a human have to approve something before side effects happen?

'Ask the end user before running' — 'This option is set to No by default.'

Is there a plan/run confidence score derived from checkable structure rather than asked of the model?

Does a correction on a finished run alter behaviour on subsequent runs without a human editing a prompt?

Thumbs-up/down in evaluations improves the GRADER ('evaluation reliability'), not the agent. Memory (preview) is per-user, private and 28-day-expiring — personalisation, not organisational learning.

Can a person list what the system has learned, see where it was applied, and retire it?

Is a team of differently-roled agents a first-class product concept?

Can Claude/ChatGPT approve or reject pending work through the MCP surface?

Can an outside system start a run over HTTP?

Event triggers exist but are en-US only and 'can only use the agent maker's credentials'.

Does the platform mint an address that starts a run, as opposed to polling a mailbox you own?

Can the customer run it in infrastructure they control?

Is there a durable process definition distinct from an individual run of it?

Can an individual tool call be gated mid-execution?

Is there a review gate on the deliverable before it is sent?

Is there version history with a way back to a previous version?

Is there a promotion path between environments, or git-backed source control?

Full Power Platform ALM — dev/test/prod sandboxes, managed solutions, Pipelines, Azure DevOps, GitHub Actions. Not solution-aware: channels, auth settings, sharing, App Insights, multistage approvals.

Can you adopt a ready-made agent persona, as opposed to a workflow template?

Can a run start on a schedule?

Can work be started from a chat platform?

Is there a machine credential, so server-to-server callers need not hold a user session?

Is there a SOC 2 Type II report?

Is the vendor ISO 27001 certified (not merely 'aligned' or engaged)?

Does each execution produce a plan artefact before the first unit of work runs?

Plans are runtime objects: there is an 'A plan completes' trigger and 'End all topics' cancels 'remaining steps the orchestrator planned'.

Can knowledge be attached to an individual agent and retrieved at run time?

Can it consume external MCP servers as tools?

Is there a per-run record of what happened, step by step?

Is there a documented EU location for the customer's stored data?

EU Data Boundary covers Copilot Studio if the tenant billing address is EU/EFTA and all environments are in-boundary.

Is there a documented way to constrain where the prompt is processed, or to bring your own model?

'If your Power Platform and Dynamics 365 environments are hosted in the EU Data Boundary, we use an Azure OpenAI endpoint in the same boundary.' BUT 'Flex routing is on by default for eligible tenants that were created after March 25, 2026' — clearing it is what 'results in all LLM inferencing occurring inside the EU Data Boundary'. Bing Search grounding is US for every region.

Is the product interface available in Czech?

'Czech (cs-CZ)' supported for the authoring canvas, generative answers and voice.

Can two versions be compared side by side?

Does it expose itself over MCP to external AI clients?

Is that MCP server protected by OAuth 2.1 with spec discovery, rather than a static bearer key?