SmartStaff.io vs Kognitos

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

Can a person read that plan before the work starts?

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?

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

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

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

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

kognitos_control_run offers pause/continue and reply_to_exception resolves a waiting exception — adjacent, but it resolves an exception rather than clearing a pre-execution gate.

Can the customer run it in infrastructure they control?

No. What exists is Remote Connections — an IPsec VPN tunnel so cloud-hosted Kognitos can reach on-prem systems. Reaching your network is not running in it.

Is the product interface available in Czech?

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

Can an individual tool call be gated mid-execution?

Exception-driven, not gate-driven. A run enters Waiting when the agent cannot resolve something; Validation exceptions mean 'the automation needs confirmation that a condition or business rule has been met'.

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

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

'Each time you publish a draft, Kognitos creates a new major version… Published versions are locked and repeatable. New runs always use the latest published version. In-progress runs continue on the version they started with.'

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

Each guidance entry records Root Cause, Guidance, creator and timestamp, plus an 'Applied to Runs' table listing every run where it fired. Searchable and deletable without affecting past runs. Best in field.

Can a run start on a schedule?

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)?

Claimed in the site FAQ alongside SOC 2 Type II, HIPAA and GDPR, with 'ISO/IEC 42001 alignment work underway in 2026'. The trust centre blocks automated access, so this is vendor-stated rather than independently verified.

Can two versions be compared side by side?

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

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

BENCHMARK. Troubleshooting Guide: 'When an exception occurs, Kognitos checks the guide for resolution instructions. If they exist, it applies them automatically. If they don't, it pauses to ask for help.' Kognitos also mines resolution patterns and suggests new guidance. Limitation: guidance is scoped per automation with no cross-automation propagation.

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?

Can an outside system start a run over HTTP?

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

'An email trigger starts a run whenever an email is received at a unique address generated by Kognitos… (e.g., automation-abc123@us-1.kognitos.com)'. Content and attachments become inputs; filters on sender, CC and subject.

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

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

Can it consume external MCP servers as tools?

Can work be started from a chat platform?

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

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