Chapter 8: The Second Permission
by inkadminThe source review opened with forty-three minutes left on the freeze.
The remaining time glowed at the edge of the sealed recorder’s display, small enough to ignore at a glance and impossible to forget once noticed. Every few seconds the digits changed without sound. No alarm accompanied them. No warning tone was necessary. The freeze held the disputed instruction, its dependent records, and the protected route in their present state. When the window closed, whatever came next would have to proceed under a different authority and a different risk.
Berth Six had ventilation again, though every breath tasted of repair foam. The restored airflow pushed the chemical smell through the room in slow, uneven currents. A temporary seal along the upper wall flexed whenever the fans increased pressure. Nothing leaked, but the faint movement kept drawing Mara’s eye.
Ilyan set a water pouch beside her without comment. He had already broken the seal and folded the drinking valve outward. Mara left it there.
Tomas Vale’s effects case waited beyond his elbow, its transfer docket still unsigned. The case had remained exactly where preservation procedure required, close enough to be visible from the review table but outside the active work surface. Its seals were intact. The unsigned docket lay beside it under a transparent restraint strip, the transfer fields blank and the custody fields current.
No one in Berth Six mentioned the case when the review began.
The protected objection arrived first.
It displaced the blank comparison field on the central display, appearing in block text beneath the review identifier.
SCOPE REVIEW WILL INDIRECTLY DISCLOSE THE ENTITLEMENT HOLDER.
The objection contained no visible signature, endpoint, or routing history. Its protection markers were valid. That much the system exposed, and nothing more.
Leth Ardin read it twice. The first time, his expression did not change. The second time, he leaned nearer to the display as if the wording might narrow under inspection.
“Names can remain screened,” he said. “Clauses cannot become secret merely because someone used them.”
The reviewer did not answer immediately. A status line showed that the objection had been received, authenticated, and preserved. The countdown continued at the recorder’s edge.
Mara kept her attention on the distinction Leth had drawn. The entitlement holder’s identity could remain protected. So could endpoint names, account references, and any route detail capable of resolving them. None of that required the review to accept an invisible grant with invisible limits. If the continuing instruction claimed authority from an entitlement, then the operative permissions had to be compared with the outputs the instruction produced. Otherwise, “protected” would function as an answer to every question.
The reviewer admitted the source instruments under sealed comparison.
The display divided into two controlled fields. One contained the continuing instruction as already preserved. The other showed authenticated grant terms after filtering. No source document could be copied from the field. No endpoint names appeared. The instruments were not opened for general inspection, and their surrounding records remained screened.
What appeared instead were permissions.
One manifest delivery.
Not a standing stream, not a repeating feed, and not a continuing entitlement to whatever later data might resemble the original manifest. One delivery.
Notices of object existence, custody ordinal, seal state, preservation state, and mount-state change.
The five categories appeared separately, each tied to the same screened entitlement and each expressed as a notice permission rather than a general access right. The language allowed the existence of an object to be reported. It allowed the object’s place in custody to be reported by ordinal. It allowed changes in the seal, preservation condition, and mount state to generate notices.
The final permission was internal object resolution for those notices.
Mara read that clause more slowly. Internal resolution meant the system could identify the correct object well enough to generate an authorized notice. It did not, by its own wording, convert the resolving value into a distributable output. The distinction was narrow only if one ignored what a private locator could do once released beyond the notice process.
The reviewer aligned the grant language with the continuing instruction’s output fields. The comparison did not expose the protected destination. It exposed only whether a claimed operation matched an authenticated permission.
Mara searched every operative clause.
She began with the manifest delivery provision, checking its definitions and incorporated limits. The permission was singular. Nothing in its operative text transformed the initial delivery into an indefinite right to retain the private locator used to assemble it.
She moved through the five notice categories. Object existence permitted a notice that an object existed. Custody ordinal permitted a notice of position within the authorized custody sequence. Seal state, preservation state, and mount-state change each allowed the corresponding condition to be communicated. The clauses differed in trigger and content, but none contained language authorizing the underlying locator to recur as an output.
She checked internal object resolution last, then checked it again against the instruction’s configured behavior.
None authorized retention or recurring transmission of the private locator.
None authorized acknowledgement-token lineage.
None authorized successor-locator mapping.
Mara expanded the operative terms without opening the protected identity fields. Every cross-reference available within scope led back to the same limited functions. The grants described what notices could be produced and why the system could resolve an object internally to produce them. They did not authorize a persistent locator channel alongside those notices. They did not authorize the history of acknowledgement tokens to be exposed. They did not authorize one private locator to be followed into another when the object’s mapping changed.
Ilyan stood on the opposite side of the table, hands clear of the controls. The amber reflection from his credential ledger had not appeared yet, but he watched the comparison as though he expected it.
Leth remained near the protected-objection field. “The authorized notices survive the comparison,” he said.
“Yes,” Mara said.
That mattered. The excess could not be treated as proof that the entire entitlement was false or that every output associated with it was forbidden. The source instruments genuinely authorized something. They simply did not authorize everything the continuing instruction had been configured to do.
The transaction summary opened beneath the clause comparison.
Its visible fields were sparse. The instruction’s establishment, output class, and current status were represented, but the activation time was absent from the ordinary summary. That absence had already complicated the sequence. Without an activation time, the continuing operation could appear to have no exposed beginning, only a present configuration and a chain of recurring effects.
The sealed comparison showed more.
The transaction summary’s missing activation time had not been universally suppressed. It still existed within the source record. Its absence from the reviewed summary was the result of filtering, not the absence of data.
A disclosure-filtering schema had been affirmatively selected when the continuing instruction was configured.
The reviewer displayed the schema selection only as a confirmed configuration fact. The schema itself had removed the activation time from the transaction summary available in the ordinary review path while preserving it elsewhere in the protected record.
The record did not identify who selected it.
Mara held the relevant fields side by side. The finding had to remain exactly as narrow as the evidence. Someone had selected a filtering schema in the configuration process. The record before them did not identify that person. It did not establish whether the selection had been deliberate concealment, routine handling, inherited configuration, or an automated consequence of some protected choice. It established the selection and its effect. Nothing further.
A separate field appeared beneath the comparison.
RECURRING OUTPUT AUTHORITY: OVERRIDE CLASS REQUIRED. REVIEWED GRANT MATCH: NONE. SIGNER: PROTECTED.




0 Comments