Payments Agent
The run is the last point at which anything is reversible. Everything after it is a recovery conversation.
A payment run contains only entries that passed every check.Eligibility is decided by what cleared at entry, not by what is due. An invoice that is still held cannot be swept into a run because the date arrived.
What it checks
On every invoice, not a sample.
- Every line in the run traces to an entry that cleared its checks
- Held items cannot enter the run because their due date passed
- Amounts in the run match the approved entry, not the original invoice
- Nothing is in the run twice under two references
What it reads, what it writes
Reads
Approved entries and their check history · The proposed payment run · Payment terms and due dates
Writes
The run, with anything ineligible excluded and the reason recorded · A traceable link from each payment back to the invoice and the checks behind it
Where it sits in the team
Agents do not work alone. Each one hands its result to the next, so a finding raised here shows up as context downstream rather than being re-derived.
Hands to it
It triggers
When it isn't sure
Where an entry is eligible but something changed since approval — an amount, a bank account, a vendor status — it leaves the run and is named, rather than being paid on the strength of an approval that no longer describes it.
Any check that runs on every transaction will meet cases it cannot settle. What makes a control trustworthy is not that it never hesitates — it is that hesitation produces a named question for a person, rather than a silent pass or a silent block.
See what this one finds in your last 90 days.
One export. No connector. We come back within a working day.
$Check your savings→