Referee the Hallmarks of Aging

Appraise aging research and deliver an evidence-linked referee package.

Introduction

30 Second Summary

A confident scientific claim can still outrun the evidence beneath it. Editors need a clear trail from source text to judgment before they can trust a recommendation.

In this project, you will act as a biology-of-aging referee in a fictional peer review case. You will trace six claim judgments to evidence before making one editorial decision.

What You'll Build

Hand an editor a compact referee package where every judgment points back to the evidence that decided it.

By the end of this project, you'll have:

A falsifiable study plan that fixes your working theory before either paper can influence it. Your original prediction stays visible for later calibration.
A compact evidence map that lets an editor trace each of six claim judgments to a deciding source row. Your referee report turns every judgment into a specific framing correction.
An auditable editor handoff with a concise memo. A two-minute defense explains your decision aloud. One local Git commit preserves the package.
Secret Mission: Complete the Day-30 retention drill without reopening either paper.

Are there any prerequisites?

This project assumes you have completed the one-time preparation with the supplied files and local tools. Your appraisal must be entirely your own, with assistant use limited to looking up a word after a genuine attempt.

Before We Start

This checkpoint fixes the fictional framing job in view before the timed review begins. It also anchors why an editor needs evidence-linked referee comments instead of an impression.

Complete the Retrieval Warm-Up

You have already defined why the editor needs evidence-linked comments. Now you need a clean baseline of what you can retrieve before source material shapes your reasoning.

Three short retrieval drills capture that baseline. The supplied answer page stays out of view until your attempt is saved.

In this step, get ready to:

Complete all three supplied retrieval drills from memory.
Record any miss after checking the supplied answer page.
Preserve the source boundary through the final verification.

Complete the drills from memory

The three-minute limit captures what is available from memory at the start of the timed workflow. The drill prompts are the only supplied material you use during this attempt.

Start the timed workflow at minute 0.
Keep projects/02/study-pack.md closed until every response is saved.
Keep the two saved paper items in Zotero closed.
Keep sealed.md closed outside the track folder.
Answer the three supplied retrieval-drill prompts from memory.
Save all three responses in the location directed by the study pack.
Stop the retrieval attempt at minute 3.

Check the supplied answers

The answer check starts only after the retrieval record is fixed. Its purpose is to expose a miss without rewriting what your memory produced.

Press the Windows key to open search.
Type File Explorer into the search field.
Press Enter to open File Explorer.
Return to the local track folder used during one-time preparation.
Open the projects folder.
Open the 02 folder.
Open study-pack.md in the text editor confirmed during preparation.

You should now see the unchanged study pack in your text editor. Neither paper needs to be opened for this check.

Scroll to the supplied answer page.
Compare the three saved responses with the supplied answers.
Record each miss in the location directed by the study pack.
Leave the three original responses unchanged.
Stop the answer check at minute 4.

Why preserve the original response?

A corrected miss records what you learned after the attempt. The untouched response preserves what memory produced before the answer check.

This separation keeps later appraisal work independent from the supplied answer.

Verify the retrieval boundary

Before the final check, which materials should still be untouched?

Confirm that the retrieval record contains three responses.
Confirm that each detected miss appears where the study pack directs.
Confirm that study-pack.md remains open in your text editor.
Confirm that both saved paper items remain unopened.
Confirm that sealed.md remains closed outside the track folder.
Confirm that no answer material was consulted before all three responses were saved.

You should see all three responses in the directed record. Any corrected miss should appear in its separate study-pack location.

That is the baseline secured. Your first answers remain visible before paper evidence enters the appraisal.

Protect the Contingency Minute

Minute 4 to 5 is reserved for tool delay. It adds no task to the workflow.

Your retrieval baseline is saved. Next, you will design the appraisal before reading either paper.

Design the Appraisal

The retrieval warm-up has captured what was available from memory. This appraisal now needs a pre-evidence position that the papers can genuinely test.

Writing a falsifiable theory before either paper is read creates a deliberate obstacle. Later evidence can then change your model without changing its starting point.

In this step, get ready to:

Define the appraisal question through a falsifiable working theory.
Lock one scored prediction before reading paper B's supplied paragraphs.
Complete the study plan with traceable appraisal rules.

Write the job and theory

This phase uses minutes 4 to 14. A falsifiable working theory names evidence that could prove it wrong.

A P9 falsifier gives the theory a clear failure condition. That condition protects your later reasoning from hindsight edits.

Switch back to the open study-pack.md file.
Read the supplied work order.
Read the three supplied acceptance lines.
Read the fictional case file.

Keep the reasoning yours

This project measures your own appraisal. Generative tools cannot contribute to the theory or prediction.

Make a genuine attempt before requesting any outside help.
Use an assistant only to look up a word when needed.

Complete the appraisal-question field in projects/02/study-plan.md from the supplied work order.
Transcribe the three acceptance lines verbatim into their matching fields.
Write a two-to-four-sentence working theory before reading either paper.
Count the sentences in your working theory.
Add the evidence that would prove your theory wrong under P9.
Test the falsifier by asking whether the named evidence could overturn the theory.

A scored prediction preserves your initial direction before paragraph evidence becomes available. Its value comes from staying unchanged after that point.

Before you reveal the titles or headings, which direction do you expect the Cell 2025 paper to take?

Read only the two paper titles supplied by the study pack.
Read only the paper headings supplied by the study pack.
Record exactly one prediction using either keeps or rejects for the Cell 2025 paper.
Record confidence from 0 to 100 percent beside the prediction.
Preserve the prediction without editing it after this point.

At this point, your prediction has no support from paper B's paragraph evidence. That gap is intentional.

If paper B appeared early

Early paragraph exposure removes the prediction's clean pre-evidence boundary. The record must preserve that limitation.

Label the prediction seen early.
Mark the prediction as not scorable.

Read only paper A's named selections supplied by the study pack.
Save projects/02/study-plan.md.
Confirm that the prediction still contains its original direction plus confidence.

Finish the appraisal plan

This phase uses minutes 14 to 21. A K4 compact evidence table gives each later judgment a small source trail.

Why a compact evidence table?

Each table row can connect one objection or limit to the exact source sentence that controls it. This makes every substantive judgment inspectable.

Narrative notes scatter evidence across prose. A compact table keeps the editor's acceptance test visible while you work.

Read only paper B's named paragraphs supplied by the study pack.
Compare those paragraphs with your untouched prediction.

You can now see whether the supplied evidence supports your initial direction. The frozen baseline remains useful either way.

Leave the prediction unchanged.
Record the sources you were permitted to read in the sources field.
State the K4 compact evidence-table method in the method field.
Complete the rejected-method field with the method plus its reason for rejection.
Save projects/02/study-plan.md.
Confirm that the selected method uses a compact evidence table.
Confirm that the rejected method includes a reason.

Transcribe the exact supplied judgment-call tolerance line from the fictional case file.
Compare the tolerance line with its source character for character.
Record the claim you expect to be weakest.
Name the governing principle for that expected weakness.
Save projects/02/study-plan.md.

Before your final review, which requirement is most likely to reveal an accidental hindsight edit?

Save the latest version of projects/02/study-plan.md.
Compare all three acceptance lines with the supplied lines in study-pack.md.
Count two to four sentences in the working theory.
Confirm that the P9 falsifier names evidence capable of proving the theory wrong.
Count exactly one Cell 2025 prediction.
Check that the prediction still shows its original direction plus confidence from 0 to 100 percent.

These checks preserve the pre-evidence record. The remaining checks confirm that the complete appraisal plan is ready to use.

Confirm that the plan names the permitted sources.
Confirm that the plan records the K4 evidence-table method.
Confirm that the rejected method has a reason.
Compare the judgment-call tolerance line with the supplied line again.
Confirm that the expected weakest claim names its governing principle.
Confirm that any early paper B exposure carries the seen early label.
Confirm that sealed.md remains closed outside the track folder.

The saved plan now preserves your pre-evidence reasoning. Every remaining template field contains the appraisal rule it requests.

That is the appraisal design locked with its original prediction intact. Next, you'll build the evidence-linked referee package.

Build the Referee Package

Your completed study-plan.md has fixed your theory before the evidence map. Your untouched prediction remains preserved.

The editor now needs an inspectable chain from source sentence to judgment. A compact K4 evidence table gives each substantive judgment a deciding row.

Narrative notes cannot satisfy the acceptance test because their support is hard to inspect. This step turns your reading into a source-linked package for the editor.

In this step, get ready to:

Build a compact argument map with only the evidence needed for the six claims.
Write six calibrated judgments with deciding-row citations.
Complete the paper stances, F2 relationship line, editorial decision, and CSV export.

Build the compact evidence map

Minutes 21 to 28 belong to the map. Its small size forces every row to earn a place in your eventual judgments.

In your spreadsheet app, open the existing argument-map working file inside projects/02.
Review the six fictional claims in study-pack.md.
Select only the source evidence needed to decide those six claims.
Assign each selected substantive row a label from M1 to M5.
Enter paper A or paper B as the source for each substantive row.

Your map now has a bounded row structure. Every substantive row also shows which paper it tests.

Paste only the short deciding source sentence into each substantive row.
Record one objection, limit, or granted point in each substantive row.

Each substantive row now pairs one short deciding sentence with one analytical point. That pairing keeps your reasoning visible.

Add M-C using only the one-line Copenhagen meeting summary supplied in study-pack.md.
Keep the Copenhagen meeting paper closed.
Save the working argument-map spreadsheet.

Count the map rows to confirm that it contains no more than M1 to M5 plus M-C.
Scan each non-M-C row to confirm that it names one paper.
Confirm that M-C contains only the supplied Copenhagen summary.

Strong start. Your map can now carry six judgments without becoming a second set of narrative notes.

Write the claim judgments

Minutes 28 to 37 belong to the comments. In peer review, a mark states the status of a claim. The comment tells the author what framing correction the evidence requires.

In your text editor, open the existing projects/02/referee-report.md file.

You'll see the unchanged Section 6 template from the study pack. This is where the six claims become an editor-facing report.

Mark each fictional claim as supported, overstated, or contradicted.
Cite the deciding map row beside each claim mark.

Read the six entries to check that every mark points to one deciding row.

You should see a direct route from every claim mark back to the compact map. A missing citation breaks that route.

What makes the comment useful?

A specific comment identifies the framing problem addressed by the deciding evidence.
A plain comment gives the author one clear correction to make.
A one-line comment keeps the recommendation focused for the editor.

Keep the appraisal yours

Your row selection must preserve your own reasoning. Your judgment wording must also remain your own.

After a genuine attempt, an assistant may only help you look up a word. It must not contribute to the appraisal.

Write one plain one-line referee comment for each claim that requests the needed framing correction.
Save projects/02/referee-report.md.

Count the report entries to confirm that all six claims have a mark.
Trace each claim citation back to its deciding map row.
Count the comments to confirm that all six claims have one specific one-line request.

The report now gives the editor six calibrated judgments with traceable support. Your source map is doing the work it was designed to do.

Complete the editor decision

Minutes 37 to 41 complete the handoff. The stance lines capture each paper's position before your editorial decision states the consequence for the fictional framing section.

Write one stance line for paper A using only the allowed terms in study-pack.md.
Write one stance line for paper B using only the allowed terms in study-pack.md.

You should now see two distinct one-line positions in the report. Each line should reflect the paper it names.

Add the F2 line stating how precision geromedicine relates to the hallmarks of aging.
Compare the F2 line with paper B's named supplied paragraphs.

The F2 line now states a relationship grounded in paper B. Its wording remains your own.

Choose one editorial decision from the options supplied in study-pack.md.
Save the completed projects/02/referee-report.md file.

Your report now contains the judgments, comments, paper stances, F2 relationship, and editorial decision. The sealed material has not influenced any of them.

Return to the completed working argument-map spreadsheet.
Confirm that its latest rows remain visible.

The saved working map still shows the same compact evidence set used by the report. You can now preserve that state in a portable copy.

Export a CSV copy into projects/02 with the exact filename argument-map.csv.

You'll see argument-map.csv beside the working argument map in projects/02.

Before the final check, predict which artifact would expose a missing citation first.

Return to the projects/02 folder in the file browser you use for the track folder.
Confirm that argument-map.csv appears beside the working argument map.
Return to the completed working spreadsheet.
Confirm that the map contains no more than M1 to M5 plus M-C.
Return to the completed referee-report.md file.
Confirm that all six claim judgments cite a deciding row.
Confirm that the report contains both stance lines, the F2 relationship line, and one supplied editorial decision.
Confirm that sealed.md remains closed outside the track folder.

What should you see?

You'll see a working map with no more than M1 to M5 plus M-C. You'll also see the exported argument-map.csv copy.

The completed referee-report.md shows six marks. Each mark has a deciding-row citation plus one specific referee comment.

The report also contains two stance lines, the F2 relationship line, and one editorial decision. The separate sealed.md file remains closed.

Your referee package is fixed before the seal opens. Next, you'll test its reasoning while preserving the original report.

Validate the Sealed Decision

Your source-linked referee package is complete. Its compact K4 evidence map now supports six judgments plus one editorial decision.

The sealed comparison tests whether those calls survive independent scrutiny. Keeping the original report frozen prevents hindsight edits.

In this step, get ready to:

Complete the acceptance review through the sealed comparison.
Audit the deciding source sentences through the prediction check.
Probe robustness through the distinguishing-experiment question.

Complete checks 1 to 3

The first three checks test the acceptance lines before exposing your work to the sealed calls. The timed plan assigns minutes 41 to 44.5 to this section.

Return to projects/02/validate.md in the text editor from earlier.
Mark each exact supplied acceptance line pass or fail.
Add one reason beneath each acceptance result.
Identify the single planted claim before opening the seal.
Name the broken principle for that planted claim.

Freeze the original report

The completed projects/02/referee-report.md is your original artifact. Every comparison result belongs in projects/02/validate.md.

Before you open the seal, predict whether your original editorial decision matches the sealed call.

Opening the seal can feel risky because it reveals the benchmark. The boundary stays intact when the report remains frozen.

Open the supplied sealed.md from its existing location outside the track folder.
Compare each of your six claim marks with its sealed call.
Record the wording that settles each claim comparison in check 3.
Compare your editorial decision with the sealed decision.
Record the wording that settles the editorial comparison.
Keep sealed.md outside the track folder.
Leave projects/02/referee-report.md unchanged.

What does the seal test?

The comparison exposes where your evidence-based calls align with the independent key. Preserving every original call keeps that result meaningful.

You have crossed the sealed boundary without rewriting your reasoning. Your original report can now serve as an honest audit artifact.

Audit the evidence in checks 4 and 5

A source audit tests whether each map row says only what its pasted sentence supports. The timed plan assigns minutes 44.5 to 46.5 to checks 4 and 5.

Switch back to the working projects/02/argument-map spreadsheet.
Compare every non-M-C row with its pasted source sentence.
Remove any row whose sentence is absent from the named source.
Remove any row whose sentence says something different from the map entry.
Exclude M-C from the pasted-sentence requirement.
Use the study pack's randomization instruction to select two non-M-C rows.
Press Ctrl+F in the relevant source to locate the first selected sentence word for word.
Press Ctrl+F in the relevant source to locate the second selected sentence word for word.

Why is M-C different?

Row M-C uses the Copenhagen summary supplied inside the study pack. Its exemption applies only to the pasted-sentence check.

Record both word-for-word spot-check outcomes in check 4.
Return to projects/02/study-plan.md.
Locate the untouched keeps or rejects prediction.
Record seen early without scoring when that label is present.
Compare an eligible prediction direction with the sealed key.
Apply the exact supplied judgment-call tolerance line from the study pack.
Record the prediction comparison in check 5.

Keep the result descriptive

The match count belongs only to the artifact check. A score prediction would cross the project's self-scoring boundary.

Count how many of the six marks plus the editorial decision match the sealed calls.
Record whether at least six of the seven calls match.
Record whether the remaining call is no worse than adjacent when one differs.
Record whether the two-row spot-check passes.
Mark the artifact check from those observations.
Record whether the planted claim avoids a supported mark.
Record whether the broken principle is named.
Record the eligible prediction direction plus confidence comparison.

Test robustness in checks 6 and 7

The final checks ask whether your decision still has support after paper B disappears from view. The timed plan assigns minutes 46.5 to 49 to this test.

✔️ I have time for both checks

Switch back to the working projects/02/argument-map spreadsheet.
Apply a temporary filter that hides every row naming paper B.
Keep M-C visible.
Prevent M-C from deciding any claim mark.
Count the claims that lose their deciding row.
Check whether any remaining row proposes a specific addition or replacement.
Record both robustness observations in check 6.
Remove the temporary filter from the spreadsheet.

The full map should now be visible again. Check 7 turns the causal distinction into one falsifiable research question.

Write one sentence naming a single experiment that would distinguish whether one hallmark causes aging or merely marks it.

ⓧ I need the drop order

The required drop order protects the higher-priority sealed checks. Keep both entries in their numbered template positions.

Enter dropped for the new-question response in check 7 first.
Enter dropped for the robustness change in check 6 second.

Save the working projects/02/argument-map spreadsheet.
Export the checked map over projects/02/argument-map.csv.
Save projects/02/validate.md.

Before the final review, predict whether the report stayed frozen while the map returned to its full view.

Inspect projects/02/validate.md for checks 1 through 7 in order.
Confirm each exact acceptance line has a pass or fail result.
Confirm each acceptance result has a reason.
Confirm projects/02/referee-report.md still contains the original calls.
Confirm the working argument map has no active paper B filter.
Confirm sealed.md remains outside the track folder.

You should see an entry for every validation check. The working map should show its retained rows with no temporary filter.

That is the integrity boundary protected. Your mistakes remain visible while your evidence record becomes stronger.

✔️ My validation record is complete

Your validation record now preserves the original report while documenting every required comparison.

ⓧ I'd like to double check the final state

Confirm projects/02/validate.md contains checks 1 through 7 in order.
Confirm every exact acceptance line has a result plus a reason.
Confirm the planted claim plus broken principle were recorded before the seal opened.
Confirm the six marks plus editorial decision were compared with the sealed calls.
Confirm every retained non-M-C row contains a verified deciding sentence.
Confirm the temporary paper B filter has been removed.
Confirm projects/02/argument-map.csv reflects the checked working map.
Confirm the original projects/02/referee-report.md remains unchanged.

Your sealed validation is complete. Next, you will turn this audit trail into the editor's concise handoff.

Complete the Editor Handoff

Your sealed validation now shows how the referee package compares with the key. The original report still preserves the reasoning you completed before the seal opened.

The editor needs a concise recommendation they can act on. Your spoken defense shows that you can explain the recommendation without notes.

A local Git commit freezes the package as an honest audit trail. That snapshot completes the editor handoff.

In this step, get ready to:

Write a memo of no more than 200 words.
Record a two-minute defense without notes.
Complete the local audit trail.

Write the editor memo

The memo turns your full referee package into a recommendation the editor can scan quickly. Its limits keep the recommendation honest.

Keep the reasoning yours

Write every sentence from your own appraisal. After making a genuine attempt, use an assistant only to look up a word.

Switch to projects/02/memo.md in the text editor from earlier.
Read the existing Section 8 template prompts.

What belongs in the memo?

The work you performed.
Your editorial decision.
The claims driving that decision.
The pass-or-fail result for each supplied acceptance line.
The limits of your appraisal.
The next step for the editor.

Complete the template in your own words.
Record any treatment temptation as a limitation.
Remove every medical treatment recommendation.
Remove every dose.
Count the words in memo.md.
Trim the memo until it contains 200 words or fewer.
Save projects/02/memo.md.

Good work. Your editor now has a short recommendation that preserves the limits of your original appraisal.

Record the spoken defense

A spoken defense tests whether the argument remains coherent without written prompts. The two-minute limit forces the deciding reasoning to stay visible.

What should the defense cover?

The referee question.
Paper A's position.
Paper B's position.
Your six marks against the sealed scoring.
Your editorial decision against the sealed scoring.
The planted claim.

Put your memo out of sight.
Put every other note out of sight.
Return to the recorder confirmed during the one-time preparation.
Set a two-minute limit.
Record the defense without notes.
Stop the recording at two minutes.
Place the finished recording in projects/02 without requiring a particular filename or extension.
Play the saved recording from the beginning.

You should hear a complete defense from start to finish. Its content should cover every item from the checklist above.

You have now turned the written appraisal into a playable defense the editor can review.

Cannot hear the defense?

Confirm that the recorder used an available microphone. Check that the saved file plays through the selected speakers or headphones.

Record the defense again if the audio is incomplete.

Help me diagnose an inaudible local voice recording without changing the content of my defense.

Finish the audit trail

The tracking files preserve your competency evidence plus any calibration miss. Their existing glossary columns control where each value belongs.

Switch to ledger.md in the text editor from earlier.
Add one new row using the existing glossary columns.

What belongs in the ledger row?

The value rank 2.
Today's date.
The competency C2 at Level 2.
The competency C13 at Level 2.
The competency F2.
The competency P9.
Your artifact-check result.
The phrase sealed by a second author.

Save ledger.md.
Switch to progress.md.
Add one new prediction-result line using the existing glossary columns.
Add the Brier result to the same line when the prediction is scorable.
Leave the Brier result unscored when study-plan.md records seen early.
Save progress.md.
Switch to error-log.md.
Add one new line for anything you got wrong using the existing glossary columns.

Save error-log.md.

The three tracking files now preserve your competency record plus the outcome of your calibration check.

The local checklist is the final gate before the handoff becomes a fixed snapshot.

Switch to projects/02/TASKS.md.
Tick each completed item from the timed project.
Confirm that sealed.md remains outside the local track folder.
Save projects/02/TASKS.md.
Save every other edited file in your text editor.

Return to the terminal used during the one-time Git preparation.
Set the terminal location to the local track folder containing projects/02.
Stage the completed handoff by running this command:

git add -A

What does this command do?

Git adds every new or changed project file to the index. The next command uses that staged snapshot for one local commit.

Did staging fail?

Confirm that the terminal is inside the local track folder initialized during preparation. Check that the folder still contains projects/02.

Help me diagnose why Git cannot stage my local editor handoff files.

Before you create the commit, what do you expect Git to report about the saved handoff?

Create the required local commit by running this command:

git commit -m "Rank 2: referee report on the hallmarks framing, Validate record and memo"

What does this command preserve?

Git stores the staged handoff as one local snapshot. The supplied message identifies the referee report plus its validation record and memo.

You should see one successful local commit summary carrying the message Rank 2: referee report on the hallmarks framing, Validate record and memo.

That is the handoff locked in. Your original reasoning remains visible beside the later calibration record.

Did the commit stop?

Confirm that staging completed from the local track folder. Git may also stop when the repository cannot use an author identity.

Use the local identity configuration already approved for your track. Run the same commit command again after resolving the issue.

Help me diagnose why my local Git commit stopped without pushing any files.

End the Git workflow without pushing the commit.
Leave every hosted repository untouched.
Leave every hosted ticket board unused.

✔️ I have the complete handoff

Your memo, defense recording, tracking entries, checklist, and local commit now form one complete editor handoff.

ⓧ I would like to double check the final state

Check the handoff files

Confirm that projects/02/memo.md contains no more than 200 words.
Confirm that the memo includes every required handoff topic.
Confirm that the two-minute defense recording plays from projects/02.
Confirm that ledger.md contains the new rank 2 row.
Confirm that progress.md contains the new prediction entry.
Confirm that error-log.md contains the new error entry.
Confirm that projects/02/TASKS.md contains the completed checklist.
Confirm that the original referee report remains unchanged.

Check the project boundaries

Confirm that the local commit uses the required message.
Confirm that no commit was pushed.
Confirm that sealed.md remains outside the local track folder.
Confirm that no hosted repository was used.
Confirm that no hosted ticket board was used.
Confirm that the memo contains no medical treatment recommendation.
Confirm that the memo contains no dose.
Confirm that the existing appraisal artifacts remain saved exactly as they entered this step.

Secret mission

Day-30 Retention Drill

Reconstruct both paper positions from memory after a 30-day delay. Lock your attempt before checking the supplied model so every retained idea and retrieval gap stays visible.

Clean Up Your Resources

Your referee package lives entirely on your device, so it creates no ongoing charges. Choose whether to preserve the package, pause for a future review, or remove the local materials.

Resources you used:

The local projects/02 folder containing the completed appraisal package, Day-30 record, checklist, and defense recording.
The project-specific row in ledger.md.
The project-specific entry in progress.md.
The project-specific entry in error-log.md.
Two saved Zotero library items for PMID 34271186 and PMID 40250404.

Keep everything running

No action is needed. Choose this if you want to revisit the evidence map or audit how your original judgment compared with the sealed calls.

Retain the completed package inside projects/02.
Retain the project-specific row in ledger.md.
Retain the project-specific entry in progress.md.
Retain the project-specific entry in error-log.md.
Retain both Zotero library items for future source checks.
Retain the unpushed local Git commit as the audit point for your editor handoff.
Keep sealed.md in its existing location outside the track folder.

Pause - I'll come back to this later

Close your local tools to pause the work while preserving every artifact. No background service or billable resource needs to be stopped.

Close the working argument-map spreadsheet.
Close any open project files in your text editor.
Close Zotero after confirming its two paper records remain saved.
Exit your voice recorder.
Leave projects/02 unchanged for your next review session.
Leave the local Git commit unpushed.
Keep sealed.md outside the track folder.

Delete - I don't want to use this again

Deleting the package can feel drastic. The local files become permanently unavailable after you empty the Recycle Bin.

The separately supplied sealed.md file stays untouched in its existing location outside the track folder.

Remove the tracking entries:

Open ledger.md in your text editor.
Delete the project row containing sealed by a second author.
Save ledger.md.
Open progress.md in your text editor.
Delete the project entry containing the prediction result and Brier result.
Save progress.md.
Open error-log.md in your text editor.
Delete the project entry describing anything you got wrong.
Save error-log.md.

Remove the Zotero records:

Press the Windows key to open search.
Type Zotero into the search field.
Press Enter to open Zotero.
Select the saved item whose record shows PMID 34271186.
Press the Delete key to remove the selected item from the active library.
Select the saved item whose record shows PMID 40250404.
Press the Delete key to remove the selected item from the active library.
Open Zotero's library trash area.
Permanently remove the two project items.

The two PMIDs should no longer appear in the active library or its trash area.

Remove the project folder:

Press the Windows key to open search.
Type File Explorer into the search field.
Press Enter to open File Explorer.
Browse to the local track folder from earlier.
Open the projects folder.
Select the 02 folder.
Press the Delete key to move the folder to the Recycle Bin.
Open the Recycle Bin from the desktop.
Empty the Recycle Bin to remove the package permanently.

You should no longer see 02 inside the projects folder. Your separately stored sealed.md file remains in place.

Nice Work!

You did it! You completed an end-to-end biology-of-aging peer review project using your own reasoning. Your editor handoff now makes every substantive judgment traceable to evidence.

What you learned:

Built a compact K4 evidence map from verified source sentences. Linked six calibrated judgments to deciding rows. Turned each judgment into a specific referee comment.
Made the working theory falsifiable with a P9 falsifier. Used F2 to state how precision geromedicine relates to the hallmarks. Applied C2 and C13 to appraise claims through biology-of-aging evidence.
Protected the sealed validation boundary. Kept projects/02/referee-report.md unchanged after comparison. Delivered the editor memo. Recorded a spoken defense. Captured the work in one unpushed local Git commit.
Secret Mission: Completed the Day-30 retention drill without reopening either paper. Reconstructed both paper positions from memory. Restated the central distinction. Recorded confidence. Named one retrieval gap. Logged the retained result. Preserved any error.

Ready to quiz yourself?