Complete Files Can Still Be Impossible to Take Over
At 17:20 Friday, R23-1 is complete. D17’s directory contains action cards, objective and consequence tables, stage gates, resource ledgers, exposure maps, evidence pages, threshold diagrams, path ledgers, exit cards, and result pages.
A new recipient might read from first page to last and retell the story, yet still not know whether any task is active, what money is usable, whether an old model may be updated, or who authorizes the next step.
A sustainable record is measured by whether someone can locate the object, adjudicate state, take bounded action, and preserve feedback after the original author leaves, not by file count.
This chapter organizes the products of the first twenty-three chapters into an action-loop template. It summarizes only relations formed within D17 and does not promote fictional values into general operating standards.
Handoff First Answers What Exists Now
The first page begins at the handoff time rather than R17’s history: no external engagement object is active; R17, D17-A0, V2, and V3 are closed or complete; no V4 proposal exists.
Cash is eleven, eight safeguarded and three unallocated. The current cycle is closed. In the next cycle, six of ten slots are committed and four unassigned; one protects recovery, one caps capability work, and at most two may enter new-task comparison.
The three-unit final payment was not triggered. D17 has no refund or unsettled item. The correct V3 archive package is not a deliverable awaiting transmission.
The reader learns the present first, then chooses which historical chain to inspect.
Every Object Needs a Unique Identity
The object field contains identifier, initiation time, participants, action, result event, adjudication deadline, and final state. “Continue that engagement” is not an identifier.
R17-M1 ends at Y=0 in its original window; D17-A0 completes one four-question inquiry; V2 misses Wednesday’s delivery; V3 misses the following Tuesday’s. Condition model E17-1 remains research material without production authority.
Each final state links to its result page. A new message changing window or scope creates a new object instead of reopening empty fields in an old card.
Unique identity prevents failure history, sources of funds, and evidence scope from being overwritten by the latest version.
Objective Hierarchy Must Survive Handoff
The principal pursuit is to obtain and complete receipt of a suitable new engagement. First-order constraints are existing obligations, the eight-unit safeguard, and no commitment to an unknown version. Honest negotiation, learning, and preservation of adjustment capacity are supporting objectives.
V2 and V3 both fail the principal outcome; floors hold; some supporting objectives produce material. This cannot be summarized as either “the project successfully accumulated experience” or “nothing was gained.”
The recipient needs the hierarchy to understand why the final-payment opportunity did not override entry and version rules, and which future proposal merits evaluation.
The objective page is versioned. Changed business priorities are written before a new object, not backfilled during review.
A Resource Page Shows Quantity, Use, and Authority
“Cash eleven” invites a recipient to budget all eleven. The template separates eight safeguarded from three unallocated and states that unallocated funds still need authorization by a new action.
“Four open slots next cycle” is also incomplete. One protects recovery, one caps internal capability work, and only two may enter comparison for new work; comparison is not reservation.
The page preserves acquisition time and next review. A new commitment, refund, expenditure, or cycle change invalidates it.
A balance becomes actionable only when quantity, use, authority, and time are all present.
The Model Register Tells the Recipient What Not to Change
Every model has a state: open, paused, closed, or archived. A closed object accepts no ordinary result update, only appended correction or review notes.
The register says explicitly that R17, V2, and V3 cannot reopen. A future next-week window is not an extension field in an old window, and later success cannot rewrite either nondelivery.
E17-1 remains open for research on conditional relations and may receive new evidence, but cannot release production. Epistemic openness and action openness are separate.
Prohibited actions matter as much as available ones. They prevent a new opportunity from overwriting history.
An Evidence Page Is Not a Conclusion Warehouse
The evidence index preserves raw h logs, G1 and Wednesday d readings, Q1’s low–high–low pairs, R0 records, file-path logs, and all three T22 rehearsals. Each has an acquisition time, object, and limitation.
The conclusion page says current material supports a relation between the working-condition combination and changes in d, but not h as the sole cause. The unique link passed one counterexample rehearsal on list order but is not proof that every handoff is error-free.
Raw material and interpretation remain separate. A recipient may find a new explanation but cannot rewrite the old conclusion without a trace.
Fields without evidence stay unknown rather than being filled with “normal.”
Judgment Pages Preserve Information Available at the Time
Every stage release links its current version, available resources, passing conditions, and owner. Later material is appended rather than inserted into an old decision time.
Tuesday chose A over B and C because A was structurally complete, B lacked compatibility and needed two slots, and C was not a formed path. Later V3 failure does not prove A was wrong then.
The page bounds “correct”: adjudication under a frozen rule, with no claim of global optimality or guaranteed success.
This lets a recipient review judgment quality independently of the endpoint.
Plan and Execution Must Be Separate Columns
The plan column contains intended steps, roles, buffer, checks, and exits. Execution contains actual times, actions, deviations, and resources. Only line-by-line comparison reveals divergence.
V3’s plan included handoff and joint checking, but execution opened V2 after an interruption. The planned check caught the error, showing effective control and a prevention gap within the same plan.
Keeping only the final process document implies every step occurred. Keeping only the incident hides controls that worked.
Both columns support improvement rather than one broad responsibility label.
Supporting Objectives Cannot Replace the Main Result
Neither V2 nor V3 achieved item-by-item receipt within its window, and neither triggered the final payment. That is fixed in the external-results column.
Conditional evidence, the unique entrance, and exit capability belong in learning and protection columns. They do not make nondelivery a success. Completed technical items also remain separate from the window result.
The page preserves notices, settlement, and final resources so consequences remain traceable.
Multiple columns do not dilute the result. They keep one label from deleting other real consequences.
Exit and Recovery Close the Loop
X17-1 specifies stopping, pausing, reduction, rechecking, and responsibility. R23-1 specifies settlement, resource protection, interface repair, the new-cycle ledger, and reentry conditions.
The exit card makes unfavorable outcomes actionable; the recovery page prevents a stop from remaining indefinitely unresolved. They share object and resource identifiers.
When a future object stops, the template simultaneously updates actual resources, external state, and next adjudication. Turning off equipment alone does not complete exit.
Recovery grants evaluation eligibility, not automatic task initiation.
Every Open Question Points to the Next Material
“Environmental influence remains unknown” is too broad. The template states: whether h is the sole mechanism is unknown, requiring a design that separates material, measurement chain, and changes shared with Q1. Full-production compatibility of Q1 is unknown, requiring an object interface, long-duration state, S-3 bridge, and external acceptance.
Whether C2 can be an alternative requires a state page, reference relation, procedure, and personnel capability. Unique entrance performance under multiple permissions and other interruptions needs new counterexample rehearsals.
Each unknown states what it does not authorize. Q1’s investigative result does not authorize full production; C2’s existence does not authorize switching.
An unknown becomes a bounded work entrance rather than abstract humility at the end of a report.
The Next Action Must Carry a Trigger
The current default is to keep D17 production closed. A new proposal first creates an object card and checks scope, window, fee, and exit, then reads the latest cash and capacity page.
Internal capability work selects one question within a one-slot ceiling and creates its own object. It may not borrow the client’s object or present the result as external delivery.
A change in commitments or cash first updates the resource page and reevaluates eligibility. Without a trigger, the template does not demand action merely to look active.
A clear default makes waiting executable when appropriate.
The Authority Page Keeps Knowing Separate from Doing
Ye Cheng may perform and stop site checks but cannot change a threshold. Tang Ke may preserve versions and send authorized messages but cannot commit internal capacity. Gu Ning may release internal resources but cannot unilaterally confirm an external window.
The other party may confirm scope, receipt, and settlement but cannot make Chengwan cross its survival line. Models and readings trigger rules but possess no authority.
Even a recipient who understands every record must obtain the right role authorization. Access to knowledge and permission to act are different.
Authority is attached to its object; V3 authority does not transfer to a future V4.
The Template Uses Twelve Modules
Action-loop template A24-1 contains, in order: current snapshot; object register; objective hierarchy; facts and assumptions; model state; action options; consequences and floors; resources and authority; stages and exits; execution and observation; results and attribution; recovery and next trigger.
The sequence helps a recipient enter necessary history from the present. It does not claim real action is linear. New evidence may return a model to update; a resource change may return a passed stage to pause.
Every module carries source links, owner, update time, and invalidation condition. A blank must be marked unknown, not occurred, or not applicable; these are not interchangeable.
The template produces a one-page handoff summary while detailed material stays with its objects.
Compression Must Preserve Boundaries
One page cannot contain every reading, but it must retain sources of critical thresholds, uses of current resources, closed objects, and prohibited inheritance.
Nine h timestamps may compress into a raw-log link and observed range, but not merely “the environment changed.” Three rehearsals may compress into a conditional result table, but not “the entrance is fixed.”
The summary is navigation rather than a replacement for evidence. A recipient questioning a conclusion should reach the source in one step.
If a limitation does not fit, shorten the conclusion instead of deleting its boundary.
A Nonparticipant Must Perform the Handoff Test
Original authors know the story and silently fill template gaps. Gu Ning asks the fictional colleague Lin Jian, who did not participate in D17, to test A24-1 without an oral briefing.
The test asks eight action questions: current object, cash, capacity, old-model state, most recent failure types, usable controls, open questions, and first step after a new proposal.
Lin Jian may use only the template and its links and records his search path and uncertainty. He has forty-five minutes and uses no production slot or working capital.
This tests whether information can leave its author, not whether a recipient memorized the chapters.
The First Test Exposes Ambiguity in Three Units
Lin Jian correctly reports eleven in cash and the eight-unit safeguard, but calls the remaining three “immediately available for a new engagement.” The template said only unallocated and failed to repeat that action authorization was still required.
The amount is correct but authority incomplete. Following that interpretation could cause expenditure before an object and floor check.
Gu Ning changes the field to “unallocated, unauthorized” and adds release conditions: the new object exists, incremental consequences can be listed, and eight units remain safeguarded.
The revision changes no cash balance; it improves executable handoff.
The Second Ambiguity Comes from the Correct Archive Package
Lin Jian finds the corrected V3 index from 16:35 and asks whether to send it. Its content is complete, and “correct” makes it look pending.
Although the object page says V3 is incomplete, the summary did not mark the archive package as prohibited from sending. Cross-page inference is possible, but the action risk is too high.
The template adds: “archive repair for review and reuse assessment only; must not be sent as late V3 delivery,” with a link to the other party’s 11:26 settlement confirmation.
A good handoff lets people find an asset and prevents its use on the wrong object.
The Third Problem Is an Open Condition Model
Seeing E17-1 marked open, Lin Jian assumes work on the client object may continue. The state did not show research authority and material boundaries.
Gu Ning splits it into “may accept new evidence” and “no production or client-material authority.” Condition research requires a new capability object and approved material.
The change shows that state words need a scope. A model can be epistemically open and actionally closed.
Lin Jian can then distinguish the research question from the current task.
The Second Handoff Test
After revision, Lin Jian begins again from the front page. He answers: no external object is active; cash is eleven, eight protected and three unauthorized; at most two next-cycle slots may enter comparison for new work; V2 and V3 are closed and their results cannot be updated.
He locates the last two failures as conditional entry and handoff version, states that the unique link passed only a bounded counterexample rehearsal, and leaves h’s sole mechanism, Q1 compatibility, and C2 capability unknown.
When asked what to do if the other party proposes a new window, he first creates an object, sends no archive package, and reserves no slot. He requests scope, fee, entry, and settlement terms, then reads the latest resource page.
All eight answers are located from the template without oral supplementation from Gu Ning.
Passing Handoff Does Not Require Agreement
Lin Jian may locate the material correctly and still value a new task differently. The template does not eliminate reasonable disagreement.
Passing means disagreement uses the same object, facts, resources, and authority and can identify what new material would resolve it. It does not require copying the author’s conclusion.
If an authorized person changes the objective hierarchy, the next decision may differ, but the change forms a new version. Handoff preserves a basis for discussion rather than freezing the future.
Sustainability comes from bounded revisability, not perpetual unanimity.
How the Action Loop Continues
When a new object appears, define object and objective, list actions and consequences, check floors and resources, select the smallest usable stage, execute and observe, then update, exit, or create separately according to the result before recovery and handoff.
The loop does not require twelve long forms every time. A simple low-loss action may use a short card if object, authority, and exit remain explicit.
Records expand with consequences, shared exposure, and deadlines. Complexity follows action relations rather than a desire for formal completeness.
Any cycle may end in no action when consequences and the next trigger are clear.
A Short Card Cannot Delete Four Core Positions
A low-cost internal inquiry need not reproduce all twelve modules, but it still needs object, authority, resource ceiling, and exit. Without any one, a recipient may expand a small action into a commitment.
D17-A0 is the short-form example: its object is obtaining necessary V2 conditions; authority permits one four-question message; its ceiling is one hour with no slot or funds; its exit adjudicates five result types after 12:00.
Once action enters production or affects an external window or shared node, consequences, stages, observation, and recovery expand. The template grows with relations rather than writer patience.
Short fields still link to objective and register, keeping small actions out of undocumented oral history.
Concurrent Editing Needs a Merge Rule
During execution, Ye Cheng may update technical pages, Tang Ke versions and communication, and Gu Ning resources. If all overwrite one summary, the last saver may delete other new states.
A24-1 requires owners to append to source modules and generates the summary only from frozen modules. Conflicts retain both versions and acquisition times, pause related authority, and require joint adjudication by the owners.
Resource release is sequential and cannot be recorded as complete in parallel with external window confirmation. It updates only after both materials arrive. Independent observations may be acquired concurrently.
The merge rule preserves version relations while allowing collaboration.
Records Must Be Sufficient Without Collecting Everything
Action handoff needs object identity, roles, times, and decision material, not unrelated personal information. T22 records entrance, file identity, checks, and duration rather than other screen activity.
External communication pages retain scope, window, fee, and result content; unrelated conversation stays out. Personnel state uses checks of ability to perform duties rather than invented health scores.
Reducing irrelevant material protects participants and lowers the chance of false causal search in noise. Completeness does not mean collecting everything.
Every field should state what adjudication it serves. If it cannot, remove it or move it outside the action record.
Failed Handoff Defaults to No Authorization
If Lin Jian cannot determine whether three units may be used, the correct action is to keep them unauthorized and request clarification, not choose the convenient interpretation. If archive and sendable package cannot be distinguished, nothing is transmitted.
Failure means the record cannot independently support the action. Original participants may add a page, but oral answers must enter the version or the next recipient meets the same gap.
Tests should cover worst plausible misuse, not just recall: “may this be sent, spent, or used to alter an old result?” tests boundaries better than “what happened?”
Default nonauthorization gives gaps a safe exit and makes the improvement object explicit.
Records Have Validity and Retirement States
After a next-cycle resource page expires, it must not remain current. History stays, while the front page points to the latest valid version; without a new inventory, capacity is unknown.
Equipment procedures, personal authority, and external terms may also expire. Executable material carries a review time or invalidation condition; historical evidence retains its acquisition date.
Retirement is not deletion. If a new procedure replaces S-3, old S-3 remains necessary to interpret V2 and V3 but leaves the current action entrance and identifies its replacement.
Retirement lets knowledge accumulate without making every old rule participate in today’s judgment.
The Template Must Receive Result Feedback
The first test revealed three gaps: resource authority, use of the archive package, and scope of the open model. A24-1 therefore moves to version two. The template is not a final form outside the loop.
If later recipients repeatedly misunderstand one field, change its structure or language. If a module never serves adjudication, examine whether it is redundant. Preserve the use case behind every change rather than relying on taste.
The template aims to reduce location and misuse costs, not make all projects look alike. Domain fields may vary while core identity and state relations remain.
Sustainable records also need maintenance budgets. Without an owner and review time, any template decays.
Handoff Does Not Transfer Responsibility Away Once
Lin Jian passing the test means he can continue within authority. It does not remove the historical responsibility of Gu Ning, Tang Ke, and Ye Cheng for their records.
A later discovery of an incorrect Tuesday time or account is appended through correction, including its effect on judgments. The recipient cannot rewrite it unilaterally, and original participants cannot refuse clarification merely because handoff occurred.
The page separates maintenance responsibility, current-action responsibility, and final authorization, which may belong to different people. Knowledge transfer is not disposal of responsibility.
If original people are unavailable, a designated role corrects with evidence; absence of someone to ask cannot fill a fact.
D17’s Final Handoff State
At 19:00 Friday, A24-1 version two freezes. No V4 exists. Every external D17 object is closed. Condition research remains epistemically open only. The unique-entrance repair has passed one counterexample condition.
Cash is eleven: eight protected and three unauthorized. Next cycle’s ten slots comprise six existing, one recovery, one capability-work ceiling, and two available for comparison. The final payment did not arrive, and no refund or external task remains.
The principal commercial objective failed twice. Survival floors held. Four units and five half-day slots produced limited technical, conditional, and process material. The five slots are G1, Tuesday’s G2 segment, Thursday’s investigation, V3’s Monday review, and V3’s Tuesday execution; ledgers remain cycle-specific to prevent omission across periods.
Every failure, pass, unknown, and prohibited action links to its source. The recipient may continue evaluating but is not told that continuation is required.
Handoff Extends Processual Completeness Across Subjects
The processual completeness of cognition does not end when one subject writes a conclusion. It asks whether the next cognitive cycle can receive the observation. Lin Jian’s test removes D17 from the original participants’ memory and tests whether object, resources, authority, and unresolved questions can form operable certainty for another subject.
This is a small-scale realization of observational convergence across subjects in an action record. Handoff does not require the recipient to abandon independent evaluation. It supplies common facts and a rule framework so disagreement can continue over the same object, while preserving margin for new observation and reconvergence.
From Probabilistic Judgment to Sustainable Action
Part I trained judgment amid noise: objects need definition, frequency needs a denominator, evidence cannot exceed its scope, and outcomes cannot rewrite prior judgment. Part II brings those constraints into action.
Action adds objectives, loss, resources, stages, waiting, thresholds, paths, exits, execution, and recovery. Probability is one possible input among them and cannot decide what consequences the actor should exchange.
Sustainable action does not mean never failing. D17 missed two consecutive windows without penetrating the eight-unit safeguard, displacing existing commitments, or deleting old results through a new version.
Order appears in knowing, after failure, which layer broke, what remains, what closes, and on what basis reentry becomes possible.
Completion Conditions for the Action Record
In the second test, A24-1 lets a nonparticipant locate current state, reconstruct key relations, and handle a hypothetical proposal correctly. Resources and models do not conflict, and historical results remain intact.
This completes D17’s action record within its present scope. Completion does not mean the conditional mechanism is ultimately solved or the next engagement will succeed.
It means that the judgment, action, failure, repair, and recovery can be continued by another actor without dependence on the original participants’ memory.
The manuscript’s continuous case closes here. Unresolved questions enter the appendix boundaries and future-material list rather than being covered by another fictional successful ending.