People & governance
Know which obligations have evidence behind them.
Agents can read what is outstanding, flag what is missing evidence and remind the accountable person. They cannot mark anything satisfied, which is the only arrangement that makes letting an agent near this useful.
What connects
Your obligation and evidence register, through its API.
What crosses the boundary
- Into the agent: obligations, owners, due dates, evidence status.
- Out of the agent: flags, and reminders to the accountable person.
- Never: marking an obligation satisfied, or altering a policy document.
The capabilities this exposes
- Read the obligation register
- Flag an obligation
Setting it up
- 1Authorise the connection.
- 2Confirm which obligations are in scope.
- 3Grant read and flag to the compliance agent.
Permissions
Read and flag. We do not request the ability to mark an obligation satisfied, change a review date or edit a policy. An agent that could close its own finding would make the register worse than not having one.
Jobs an agent can own once this is connected
- Flag obligations with missing evidence
- Remind the accountable person before a date lapses
- Prepare an evidence summary for a review
Where it stops
An agent cannot mark anything satisfied or resolve its own flag. It also cannot declare compliance with anything — no agent output on this platform says you are compliant, because no software can.
People & governance
Others in this category.
Give AI a job.
Build an agent, choose what it can access, define what it may do, and deploy it to work safely across your business.
No code required · Full SDKs and API · Stop any agent instantly · [email protected]