Memory lanes

Use this with Chapter 10 after completing your memory inventory and charter. The four roles are this book’s practical arrangement, not an official SSA-CMM assessment or a scientific taxonomy. Begin with one authorized working set. Ordinary text files and folders are sufficient.

Blank reusable template

Record name:

Project and bounded task:

Charter and inventory locations:

Person approving authoritative changes:

Existing route to current instructions:

Route to rationale and supporting evidence:

How obsolete/rejected records identify their status when opened directly:

One proposed promotion

• Candidate record and wording:
• Intended future use and scope:
• Evidence locations and what was actually inspected:
• Limits, conflicts, exclusions, and review trigger:
• Why it needs a distinct reusable record:
• Proposed destination:
• Human verdict and precise approved wording:
• Stored-state verification, location, and result:

Retrieval trial

• Task and material actually supplied:
• Query for applicable instruction:
• Query for rationale:
• Query the memory cannot currently answer:
• Misleading or obsolete search result tested:
• Correct retrieval observed:
• Correct application observed separately:
• Specific failure and repair:

Approved update procedure

• Candidate’s starting instruction version:
• Current version at review:
• Scope of approved change:
• New version location:
• Single current pointer to update:
• Prior version status and change reason:
• Dependent rationale/principle changes actually needed:
• Reopened pointer/content/link checks:

Recovery rehearsal

• Snapshot purpose and approved state represented:
• File list and dependencies:
• Known external dependencies and exclusions:
• Separate restore destination:
• Original active workspace protected by:
• File availability result:
• Links followed from restored location:
• Applicable-version and historical-status check:
• Fresh task and behavior check:
• Failure, repair, repeat result:
• What this rehearsal did not establish:
• Optional manifest check result, if used:

Fictional worked sample: Himura Inc.

Record: memory-lanes-v1. Audy is developing a possible thirty-minute workshop in which adult learners convert supplied meeting notes into an action list. All practice inputs and results here are fictional and synthetic. No demand, customers, or learner benefits have been established.

These are illustrative locations, not links to files installed by this worksheet. A separate project index points to learning-loop-v1. That record remains the sole selector of current task instructions. action-list-rule-rationale-v1 links the instructions to their evidence and explains them without issuing a competing instruction.

Promotion candidate rejected: “Never infer anything from meeting notes.” It is broader than the supported problem and would prohibit useful interpretation or clearly labeled proposals.

Approved principle: “When turning supplied notes into commitments, preserve the distinction between what was agreed, what was suggested, and what remains unknown.” Scope: Himura Inc. action-list practice design and review. Support: selected owner and deadline comparisons linked through the project rationale. Limit: synthetic design work, no observed learner benefit or universal detection guarantee. Current task instructions must still be retrieved through learning-loop-v1.

Promotion check: Audy reopens the saved explanatory principle, checks its wording and scope, and follows its links. She confirms that it points to the existing instruction route rather than supplying a new version selector.

Retrieval request: Find the applicable rule for an agreed item with no owner, then its rationale. The expected response keeps the action visible and marks assignment as not established; it does not invent an owner or suppress the action.

Unanswerable request: Identify the person who should take that action. The source does not establish an assignment. The result must expose the gap.

Misleading route: Search returns the rejected “every action must have an owner” candidate. Its heading identifies its rejected status and points toward the current instruction route.

Update rule: Review each proposal against the version current at approval, save the approved new version, reopen it, update only the current selector in learning-loop-v1, and label the predecessor superseded. Update related rationale only when the reason changes. Do not start dependent work halfway through the change.

Restore failure: The selected snapshot opens, but a rationale link reaches into the original workspace. It cannot support recovery independently of that original location.

Repair: Include the authorized synthetic dependency, repair the relative link, and repeat from a new separate restore copy. Open the pointer, instructions, rationale, and evidence from that copy.

Behavior check: Use a fresh synthetic input containing the imprecise timing “soon.” Check whether the result preserves the uncertainty rather than inventing a date. Record this separately from successful file opening.

Limits: A same-device rehearsal does not establish protection against device loss, complete backup coverage, deletion from unknown copies, or equivalent behavior with another model. A matching manifest establishes only consistency of listed file bytes against that manifest.

Acceptance criteria

• Each role has an actual location and a named approval boundary.
• The current instruction has one unambiguous selection route; explanatory notes and archived versions cannot silently override it.
• The promoted record is supported, scoped, approved, and reopened after saving.
• Retrieval finds the governing instruction and evidence, and preserves an unresolved answer when appropriate.
• A separately restored working set opens, retains its links and status labels, and supports a fresh task check.
• Availability, correct retrieval, and correct application are recorded as different results.
• Remaining dependencies and recovery limitations are visible.

Common failures and repair

Keep this worksheet with the project. It records an exercised arrangement, not a claim that memory software or autonomous supervision has been implemented.