The tenant-association hall had eleven address cards taped to the wall.
Salma, a community paralegal, wrote one rule above them:
**Speak only for your own case. Do not bring another person’s name without consent.**
Each participant placed timeline cards: notice, appeal, physical removal, supplier lot, online listing—only where they could verify the step.
Sofia placed four cards for the wardrobe.
Another tenant had a removal notice but no verified online item, so Musa kept that complaint outside the resale sample.
A third believed a chair was hers but could not produce a second identifying marker. Her claim remained open, not verified.
“Why is Sofia believed and I am not?” she asked.
“The standard is the same,” Salma said. “Her item has multiple corroborating markers.”
Musa knew the answer hurt, but the threshold made the stronger cases harder to dismiss.
Of the eleven addresses, seven had at least one item link. Several had documented appeals still pending when property was removed.
Listing times mattered too. Some goods appeared online within twenty-four to seventy-two hours of clearance.
Musa wrote the finding carefully: *In reviewed cases, resale timing precedes documented resolution of several appeals.*
He would not write *stolen-property network*.
People in the room wanted SokoPili to name Kifutio publicly and expose seller usernames.
Musa refused.
“Sellers have invoices. If we turn downstream buyers into enemies of owners, the upstream chain disappears from view.”
Sofia surprised the room by supporting him.
“Mando did not know.”
That changed the temperature.
Salma showed original tenant inventories only when participants allowed it. Musa looked at relevant entries, recorded references, and left the originals with owners.
Not everyone consented to take the complaint into the platform or regulator process. Those cases remained with the community association and were excluded from the operational sample.
Fewer cases, stronger consent.
Asha joined by video.
Musa proposed a targeted freeze only on new listings sourced from the affected supplier chain. Seller accounts would remain active. Sellers with alternative provenance could be whitelisted.
Legal approved.
The freeze went live while the participants were still in the room.
SokoPili’s gross merchandise value immediately fell in the affected category. Seller calls surged.
“This is containment,” Salma reminded the room. “Not restitution.”
Musa agreed. No property had yet been returned. No eviction decision had been reversed. The platform had simply stopped allowing new disputed supplier goods to move deeper into resale.
A lawyer’s letter reached SokoPili that evening.
Kifutio claimed the hold was interfering with performance under a city redevelopment tender and causing commercial damage.
The wardrobe case had now reached city procurement.
At the end of the workshop, Salma asked each participant whether their case could be shared beyond the association. Two said no. Musa removed them from the regulatory sample immediately. Sofia asked whether a smaller number would make the case look weak. “Maybe,” Musa said. “But we do not borrow people’s stories to make our numbers bigger.” Salma created a separate list for community-only assistance. That choice reduced the sample but clarified consent. When the supplier hold went live, every case used to justify it had a documented path into the process rather than merely being present in an angry room.
Salma closed the workshop by explaining what would happen next. The association would continue helping cases that had not entered the regulatory sample. SokoPili would maintain the targeted hold. The ombuds route would receive only consented references. No one in the room was promised a return date for missing property. Musa watched people leave with less certainty than they wanted but more clarity about the process. The community evidence had changed a live marketplace control without becoming a public list of alleged offenders. That was a fragile success worth preserving.
The association kept a copy of the consented timeline summary, while SokoPili stored only the references it needed. The same evidence could therefore be checked without centralizing every personal document.
Every hour of the freeze had acquired a price.