Muthoni Njeri did not hug Baraka when he arrived at Kiamwangi Cooperative.
She gave him tea, pointed to a chair and asked, “Are you here as my nephew or as an exchange man?”
“Today both may be inconvenient.”
“Then choose the one that works.”
At the wet mill, Njeri Kamau—known to everyone as Nje—was cleaning the pulper.
“Have they promoted me to witness?” she asked.
“No.”
“Good. I am an operator.”
That was exactly why Baraka wanted her.
The formal processing ledger sat in the chairman's office. Nje began with the chalkboard by the tanks: erased marks, corrections, arrows, start times and wash times.
“This is the memory of the shift,” she said. “The ledger is written later.”
Baraka asked her to reconstruct an undisputed batch first. He wanted a baseline for how the mill actually recorded work. Only then did they move to the premium batch Kiamwangi had been told was rejected.
Tank B-4 had processed that coffee. Its timing matched, almost exactly, the processing window declared for the Kifaru sample.
Muthoni said, “So it was ours.”
“Not yet,” Baraka replied.
She glared at him. “Your tongue is brave in Nairobi and afraid here.”
“If I call it ours without a chain, Victor will destroy the sentence in one minute.”
Nje opened the old shift sheets. The formal ledger said the rejected batch had moved into a lower-grade pool. The shift sheet carried another outbound code, one that did not appear on the payout summary.
Chairman Karanja Githinji entered while Baraka was photographing the board.
“Stop.”
Baraka put the phone down. “This belongs to the cooperative. If you say no, I stop.”
Karanja accused them of endangering an exporter contract that had already funded an advance.
“Advance money does not answer where our coffee went,” Muthoni said.
Baraka proposed a local custody method instead of taking originals: the board photograph, shift sheets and ledger copies would remain under cooperative control, opened by two people, with access recorded.
Karanja complained that a cooperative could not become a forensic laboratory.
“We are not building a lab,” Baraka said. “We are making sure one person's memory is not the only authority.”
Nje compared it to shift work. One operator could make a mistake; the tank board and the next shift created a second memory. That explanation persuaded Karanja more than Baraka's audit language.
They sealed copies separately and distributed custody between the secretary and chairman. Muthoni did not like giving Karanja control after his defense of the contract. Baraka insisted accountability could not mean choosing only people their side trusted.
Then Nje noticed the ghost of an older chalk mark under a newly erased line.
She copied the code.
“This was another premium lot we lost two weeks earlier.”
Its prefix matched the Kifaru route.
Karanja demanded to know whether the old chalk mark could have been copied from a later office sheet. Nje demonstrated the normal order of work: cherries entered, tanks were assigned, the board was written during the shift, and only afterward did the clerk prepare the clean ledger. The board was not sacred evidence; it was simply closer in time to the work. Baraka asked the secretary to preserve both versions instead of declaring one automatically true. That kept the question technical: which record best matched tank use, shift timing and the outbound batch that followed.
The dispute was no longer one batch sitting badly in a ledger. The same pathway had left a second mark.