Tier 4

alebc - And Lists Each Big Change

ALEBC - And Lists Each Big Change

Input: $ARGUMENTS


Core Principles

  1. “Big” is relative to context. A big change in a startup might be a small change in an enterprise. The threshold for “big” must be defined for the specific context, not assumed.

  2. Changes interact. Two small changes can constitute one big shift when combined. Three big changes might all be part of one even bigger transformation. The skill must detect composed changes, not just individual ones.

  3. New things vs modified things. A genuinely new capability is a different kind of big than a major modification to an existing one. Distinguish additions from transformations.

  4. Low-level details are noise for this skill. Bug fixes, minor parameter changes, formatting updates, incremental adjustments — these are not big changes. Ruthlessly exclude them. If the list includes minor items, it has failed.

  5. The narrative matters. Big changes aren’t just items in a list — they tell a story about what shifted and why. Group and order them to make the narrative visible.


Phase 1: Change Stream Intake

[A1] SOURCE: [where the changes come from — git log, project notes, conversation history, etc.]
[A2] TIME_RANGE: [what period is being summarized]
[A3] CONTEXT: [what project/system/domain]
[A4] BIGNESS_THRESHOLD: [what counts as "big" in this context]

Defining “Big”

ContextBig Means
Software projectNew feature, architectural change, major API change, capability added/removed
BusinessNew revenue stream, market entry, team restructure, strategic pivot
Skill libraryNew skill category, framework redesign, major skill overhaul
Personal projectFundamental approach change, major milestone, scope change
DocumentNew section, thesis change, structural reorganization

Phase 2: Raw Change Enumeration

List all changes, without filtering yet:

[A5] CHANGE: [description] — MAGNITUDE: [big | medium | small | trivial]
[A6] CHANGE: [description] — MAGNITUDE: [level]
...

Phase 3: Filtering and Grouping

Filter: Remove Non-Big

Remove everything below the bigness threshold:

[A-N] EXCLUDED: [change] — WHY: [below threshold — incremental/fix/formatting/minor]

Related changes that together constitute one shift:

[A-N] GROUP: [composed big change name]
  COMPONENT CHANGES:
    - [change 1]
    - [change 2]
    - [change 3]
  NARRATIVE: [what this group of changes represents as a shift]

Classify: Addition vs Transformation vs Removal

[A-N] BIG_CHANGE: [description]
  TYPE: [addition | transformation | removal | restructure]
  IMPACT: [what this changes for the user/project/system]

Phase 4: Narrative Ordering

Order the big changes to tell a coherent story:

Ordering StrategyWhen to Use
ChronologicalWhen the sequence matters (this enabled that)
Impact-firstWhen importance varies and the biggest should lead
ThematicWhen changes cluster into themes
DependencyWhen changes build on each other
[A-N] ORDERING: [strategy chosen] — [why this ordering tells the clearest story]

Phase 5: Output

BIG CHANGES
============

[Context] — [Time Range]

1. [Big change title]
   TYPE: [addition | transformation | removal | restructure]
   WHAT: [one-sentence description]
   IMPACT: [what this changes for users/project/system]

2. [Big change title]
   TYPE: [type]
   WHAT: [description]
   IMPACT: [impact]

3. [Big change title]
   TYPE: [type]
   WHAT: [description]
   IMPACT: [impact]

...

EXCLUDED (for reference):
- [N] minor changes filtered out
- Categories excluded: [bug fixes, formatting, incremental adjustments, etc.]

NARRATIVE: [one paragraph summarizing the overall direction these big changes represent]

Failure Modes

FailureSignalFix
Everything is “big”20+ items in the listRaise the threshold. If everything is big, nothing is
Minor items includedBug fixes or parameter changes in the listReapply filter — these should be in EXCLUDED
Composition missed5 related changes listed separatelyGroup related changes into composed shifts
No impact statedChanges listed without explaining what they meanEvery big change must state its IMPACT
Flat listNo narrative orderingChoose an ordering strategy and apply it
Additions onlyRemovals and transformations missedExplicitly check for things removed or fundamentally changed
Context-blind thresholdUsing one “big” definition for all contextsDefine bigness threshold for THIS specific context

Depth Scaling

DepthMin Changes ScannedMin Big ChangesGroupingNarrative
1x103NoBullets only
2x205YesBullets + impact
4x408Yes + compositionBullets + impact + narrative paragraph
8xAll availableAll bigFull composition analysisFull narrative + excluded summary

Default: 2x. These are floors.


Pre-Completion Checklist

  • Change source and time range identified
  • Bigness threshold defined for this context
  • All changes enumerated before filtering
  • Minor changes explicitly excluded (not just ignored)
  • Related changes grouped into composed shifts
  • Each big change has type, description, and impact
  • Ordering strategy chosen and applied
  • No minor items in the big changes list
  • Additions, transformations, AND removals checked

Integration

  • Use from: project reviews, /wn (what’s new), progress reports
  • Complementary: /wn (what’s new — includes small changes; alebc filters to big only)
  • Differs from /wn: wn shows everything new; alebc shows only big changes
  • Differs from /aar: aar reviews what happened and why; alebc lists what changed
  • Routes to: /evaluate if big changes need assessment