CHAPTER 10:
LEAKS AND HEARINGS
The Node 6A files arrived at 02:13 on August 28.
Asha Dwivedi was awake because Beijing had released another set of Gemini images and Moscow had contradicted the Chinese statement about crew selection within forty minutes. Her apartment table held three empty coffee cups, a notebook, and a secure terminal provided by the Continental Press Service after someone tried to access her private messages.
The incoming package had no greeting.
Forty-seven files.
Test data. Engineering minutes. Draft dispositions. A video recording from Lunar Gate. An incident-review questionnaire completed by Mason Kellerman but not yet submitted. The sender had included file hashes and a short instruction.
Verify before publishing.
Asha disconnected the terminal from the network.
She opened nothing for the first five minutes.
Files could be planted. Metadata could be constructed. A real document inserted among false ones could make the whole package appear authentic until publication, when an agency would discredit everything by identifying the fabrication. The timing created another problem. Gemini had been revealed days earlier. China and Russia were arguing that the Perseus institutions had failed and that another leadership structure had to act. A leak damaging NASA now served that argument whether the documents were true or not.
Truth did not arrive without a sponsor.
Sometimes the sponsor mattered less than the truth.
Sometimes it changed what the truth could safely mean.
Asha copied the package to an isolated machine and began with the least dramatic file: an equipment inventory attached to the March test.
Part numbers. Technician initials. Sensor serial numbers. Temporary cable assignments. The kind of information nobody invented for a headline because nobody outside the work knew which details mattered.
She compared the format with public Lunar Gate engineering records. The fonts matched. The date structure matched. The document identifier followed the correct internal sequence. One sensor serial number appeared in a procurement notice issued two years earlier.
She moved to the test video.
The recording began before the meeting. Empty chairs. A wall display showing Node 6A temperature traces. Voices outside the camera range discussing food and an unavailable conference room.
Then people entered.
Mason Kellerman sat near the center. Chie Nakamura took a place across from him. A contractor representative arranged three tablets in a row and spent the first minute trying to connect one to the room display.
Nothing looked staged.
Asha watched the full recording without taking notes.
The sensor divergence was real. So was the replacement. So was the repeat test inside the qualified range. The board did not ignore a red alarm and send a broken ship to Mars. It examined a condition that had become less severe after work, argued over what remained unexplained, and decided the residual risk did not justify missing the launch window.
Nine voted to accept.
Two voted against.
The worst sentence was not spoken by a villain.
It came from a mission-assurance representative near the end.
“We have reduced the uncertainty to the point where program consequences must be considered alongside technical consequences.”
Nobody objected to the sentence because it described what review boards did.
Asha replayed it.
Program consequences.
Crew exposure schedules. Cargo degradation. Lunar Gate capacity. Funding agreements. Mars geometry. Political support.
June.
All of it entered the room without anyone saying the mission mattered more than the people aboard.
They did not need to say it.
She opened Chie Nakamura’s original recommendation.
REMOVE NODE 6A FOR DESTRUCTIVE AND LOAD-EQUIVALENT INSPECTION BEFORE FLIGHT CERTIFICATION.
Below it, Chie had listed the schedule impact and acknowledged that no modeled case produced immediate vehicle loss.
The next file showed the accepted wording.
RESIDUAL SENSOR DISPERSION REMAINS WITHIN QUALIFIED RANGE.
Both documents were true.
Placed side by side, they looked like confession.
Asha called her editor at 02:57.
Samuel Reed answered on the fourth ring.
“Someone dead?”
“No.”
“Then why are you calling like someone is?”
“I have the Node 6A archive.”
He became fully awake.
“How much?”
“Enough to ruin several lives if we handle it badly.”
“Authentic?”
“Probably. I need systems, legal, video forensics and someone who understands certification boards.”
“I’ll open the office.”
“No shared network. The package stays isolated.”
“Who sent it?”
“No attribution.”
“Government?”
“Maybe.”
“Gemini side?”
“Maybe.”
“Family coalition?”
“They did not have these files yesterday.”
“You asked?”
“I know what they have.”
Samuel was silent for a moment.
“What does it prove?”
“That there was a real concern, a recommendation to remove the assembly, a nine-to-two acceptance vote and explicit discussion of the June window.”
“Does it prove Node 6A took the ship?”
“No.”
“Does it show the node misbehaved before contact ended?”
“Yes, inside the accepted range.”
“That will not survive publication.”
“What won’t?”
“Inside the accepted range.”
Asha looked at the temperature trace from Perseus’s final telemetry. The split had peaked at six-tenths of a degree and begun to fall.
“It has to.”
“You can print it. I’m telling you what people will remember.”
“They knew and launched anyway.”
“That is what they’ll remember.”
“That is also partly true.”
“Partly true is how most damage gets done.”
Asha closed the telemetry display.
“Get the team in.”
By seven in the morning, the Continental Press Service had authenticated nineteen of the forty-seven files.
The signatures in the board record matched known certification keys. The test video contained continuous camera and audio data with no detectable edits. The Lunar Gate clock aligned with independent station records. Internal document references led to public procurement and launch records where they should.
NASA received Asha’s questions at 07:18.
She gave the agency four hours.
The response arrived in nine minutes.
The documents appear to include authentic Project Perseus engineering material obtained and distributed without authorization. Technical records must be interpreted within the full certification context. No evidence establishes Node 6A as causal in the disappearance of Perseus. Unauthorized disclosure risks compromising an active international investigation.
Asha read the statement to Samuel.
“They authenticated it for us,” he said.
“They authenticated some of it.”
“They said authentic material.”
“They also said appears.”
“Good enough for television.”
“We are not television.”
“No. Television will use us as permission.”
The engineering consultant joined from Colorado. Her name was Leah Moreno. She had spent seventeen years in launch-system safety and refused to speak on record until she had seen the complete test sequence.
Asha showed her the first divergence, replacement, repeat and final board disposition.
Leah watched without interrupting.
When the video ended, Samuel asked, “Would you have launched?”
“That is not the right question.”
“Everybody keeps saying that.”
“Because you are asking one person to stand outside a program and pretend the risk had one owner.”
“Would you have voted to accept?”
Leah considered it.
“With the data shown here, I would have asked for another instrumented transition.”
“That delays launch.”
“Yes.”
“And if the instrumentation required opening the loop?”
“Then I might have voted to accept.”
Samuel looked toward Asha. “That’s useful.”
“No,” Leah said. “It is accurate.”
“What is the difference?”
“Useful means you already know where you want it to go.”
Samuel sat back.
Leah brought up the final telemetry.
“This value does not support a catastrophic thermal event. The disagreement is small. Flow and pressure are normal. It is declining.”
“Could the sensors have missed the failure?” Asha asked.
“Yes.”
“Could a failure occur between frames?”
“Yes.”
“Could Node 6A destroy every communication channel without a warning?”
“Not through any path shown in these records.”
“Could it destroy the ship?”
“If you allow enough unobserved failures, almost any system can become fatal.”
“That sounds evasive.”
“It is the opposite. The files show a governance problem. They do not show the cause of the disappearance.”
Samuel pointed at the original recommendation.
“An engineer wanted the assembly removed.”
“Correct.”
“And the board refused because of schedule.”
“No. The board accepted a residual risk while considering schedule among the consequences.”
“That is the same fact in better clothes.”
“It is not. If you write that schedule overruled safety, you are adding a conclusion the record does not support.”
Asha looked at Leah.
“What conclusion does it support?”
“That the launch window changed the burden of proof. Delay had measurable costs. Remaining uncertainty did not. That makes people more willing to accept uncertainty.”
Samuel wrote the sentence down.
Leah saw him.
“Do not remove the next part.”
“What next part?”
“They may still have made a reasonable decision.”
Nobody spoke.
Asha returned to the incident questionnaire.
Mason had selected YES beside the question asking whether programmatic or schedule pressure affected the acceptance decision.
His explanation followed.
The June 2033 departure window was a material factor in the review. It affected the consequences assigned to delay. It did not change the recorded test results. I cannot determine whether the board would have reached the same decision without that pressure.
The answer was clear enough for a headline and careful enough to resist one.
“Has this been submitted?” Asha asked.
The document history showed SAVE DRAFT. No submission stamp.
Samuel shook his head. “We cannot publish an unfinished questionnaire as a formal admission.”
“We can report that he wrote it.”
“He can say he was thinking through the answer.”
“He was.”
“And that will look like retraction.”
Asha read the file properties. It had been saved two days before the leak and reopened three times. No later revision existed in the package.
She sent Mason a request for comment.
His office replied that all incident-review personnel had been directed to refer questions to NASA communications.
Twenty-three minutes later, a second message arrived from a private address.
Do not quote this. The answer is mine. The files do not prove causation.
Asha typed a response.
Will you state that on record?
The reply took six minutes.
I have been instructed not to discuss protected review material.
She wrote again.
Did schedule pressure affect the vote?
No answer came.
At 10:40, a congressional staffer called.
The Special Committee on the Perseus Emergency had scheduled its first public hearing for September 1. Mason Kellerman, the NASA administrator, Lunar Gate certification officials and representatives of the families coalition had been invited.
“Invited?” Asha asked.
“For now.”
“Will Nakamura testify?”
“Not in the first panel.”
“Why?”
“No comment.”
“Did your committee receive the leaked archive?”
“No comment.”
“Did the leak come from the committee?”
The staffer laughed once.
“You know I cannot answer that.”
“I know you will deny it if the answer is no.”
The line went quiet.
Asha continued.
“Does Representative Park know I have the files?”
“Representative Park knows several news organizations have obtained engineering records.”
“Several?”
The staffer ended the call.
Asha looked at Samuel.
“We are not the only ones.”
“Then we publish before someone makes it worse.”
The first story went live at 12:01.
Its headline was revised eleven times.
PROJECT PERSEUS ACCEPTED UNRESOLVED THERMAL CONDITION BEFORE LAUNCH.
The opening paragraph stated that the records did not establish a cause for the disappearance.
The second described the three-point-one-degree test divergence.
The third described the removal recommendation.
The fourth described the nine-to-two vote and June departure pressure.
By the time readers reached the final-telemetry analysis explaining that Node 6A remained inside its accepted range, the headline had crossed the world.
Leave a comment for the author
Your comment is private when submitted. NDJ Productions may select comments for public display.