Tier 4

siycftr - See If You Can Find The Rest

SIYCFTR - See If You Can Find The Rest

Input: $ARGUMENTS

Core Principles

  1. Omission is not always accidental. Some items are left out intentionally (out of scope, deliberately excluded). Others are forgotten. The skill must distinguish deliberate omission from oversight. Only add what was genuinely meant but missing.

  2. Pattern completion is unreliable. “Apple, banana, cherry…” might continue with “date” (alphabetical) or “dragonfruit” (tropical) or stop at three (the list was complete). The continuation rule must be verified, not assumed.

  3. The stated items define the pattern. The items explicitly listed constrain what “the rest” could be. Three technical requirements imply more technical requirements, not marketing goals. The category of the missing items is set by what’s present.

  4. Completeness has a standard. “Did I find the rest?” requires a definition of complete. In some domains, MECE is the standard. In others, practical coverage is enough. The completeness criterion must be explicit.

  5. Better to surface-and-flag than to silently add. When uncertain whether an item was omitted or excluded, present it with a flag rather than either adding it silently or dropping it. Let the user decide.

Phase 1: Input Analysis

[R1] INPUT_TEXT: [the text being scanned, quoted or summarized]
[R2] EXPLICIT_ITEMS: [list of items explicitly stated]
[R3] ITEM_COUNT: [N items stated]
[R4] APPARENT_CATEGORY: [what kind of items are these?]
[R5] APPARENT_PATTERN: [is there an ordering, grouping, or selection principle?]

Phase 2: Pattern Identification

Identify the rule governing what’s included:

[R6] CONTINUATION_RULE: [what rule would generate these items?]
[R7] RULE_CONFIDENCE: [high | medium | low]
[R8] ALTERNATIVE_RULES: [what other rules could explain the same items?]

If RULE_CONFIDENCE is low or multiple ALTERNATIVE_RULES exist:

[R9] AMBIGUITY: [the pattern is ambiguous — multiple completions are possible]
[R10] DISAMBIGUATION: [what would clarify which pattern is intended]

Pattern Types

TypeExampleCompletion Method
Enumeration”Requirements: A, B, C…”Check if more items exist in the category
Taxonomy”Frontend, backend…”Check for missing levels/branches
Process”Step 1, Step 2…”Check for missing steps in sequence
Criteria”Must be fast, must be cheap…”Check for standard criteria in the domain
Stakeholders”Users, admins…”Check for missing roles/parties
Dimensions”Speed, cost…”Check for missing evaluation dimensions
MECE categories”Buy, build, partner…”Check for exhaustive coverage

Phase 3: Gap Identification

For each potential missing item:

[R-N] CANDIDATE: [potential missing item]
  EVIDENCE: [why this seems implied — what pattern suggests it]
  OMISSION_TYPE: [likely_forgotten | possibly_excluded | uncertain]
  IMPORTANCE: [high | medium | low]

Omission Type Classification

TypeSignalTreatment
Likely forgottenPattern clearly implies it; no reason to excludeAdd with confidence
Possibly excludedPattern implies it but it might be deliberately out of scopeFlag for user review
UncertainCould be implied or could be irrelevantPresent as “possible addition”

Phase 4: Completeness Check

[R-N] COMPLETENESS_STANDARD: [MECE | practical coverage | domain standard | N-of-M enumeration]
[R-N] COMPLETENESS_ASSESSMENT:
  BEFORE: [N items — coverage level]
  AFTER (with candidates): [N items — coverage level]
  REMAINING_GAPS: [known gaps that couldn't be filled, if any]

Phase 5: Output

COMPLETION ANALYSIS
===================

INPUT: [text or list being completed]
EXPLICIT ITEMS: [N]
PATTERN: [continuation rule — confidence level]

LIKELY MISSING (high confidence — probably forgotten):
  1. [item] — EVIDENCE: [why implied]
  2. [item] — EVIDENCE: [why implied]

POSSIBLY MISSING (medium confidence — might be excluded):
  3. [item] — EVIDENCE: [why might be implied] — CAUTION: [why might be excluded]

UNCERTAIN (low confidence — for user review):
  4. [item] — EVIDENCE: [weak signal]

COMPLETENESS:
  BEFORE: [N items — X% coverage]
  AFTER: [N items — Y% coverage]
  STANDARD: [what completeness means here]
  REMAINING GAPS: [if any]

READY FOR:
- /mv — to validate MECE completeness
- /se — to enumerate more systematically if gaps remain
- /dd — to discover missing dimensions

Failure Modes

FailureSignalFix
Invention disguised as completion”Missing items” that aren’t implied by anythingEvery candidate must have EVIDENCE tied to the pattern
Wrong patternCompletion follows rule that doesn’t match stated itemsCheck alternative rules; verify against existing items
Deliberate exclusion overriddenAdding items the user intentionally left outFlag as “possibly excluded” instead of adding
Completeness not defined”Found more items” without knowing what complete meansState the completeness standard explicitly
Category driftMissing items from a different category than explicit itemsMissing items must match the apparent category
OvercountingAdding items that are sub-items of existing itemsCheck for containment — is this already covered?

Depth Scaling

DepthMin Explicit AnalyzedMin CandidatesPattern Alternatives CheckedCompleteness Assessment
1xAll21Binary (complete/incomplete)
2xAll42With coverage percentage
4xAll74With MECE check
8xAll12All plausibleWith remaining gaps mapped

Default: 2x. These are floors.

Pre-Completion Checklist

  • Explicit items identified and categorized
  • Continuation rule identified with confidence level
  • Alternative rules considered
  • Each candidate has evidence (not invented)
  • Omission type classified (forgotten vs excluded vs uncertain)
  • Completeness standard stated
  • Before/after coverage assessed
  • No category drift in candidates

Integration

  • Complementary: /mv (MECE validation), /se (systematic enumeration), /dd (dimension discovery)
  • Differs from /se: se enumerates from scratch; siycftr completes an existing partial list
  • Differs from /etc: etc expands pattern tails; siycftr finds omissions anywhere in a text
  • Differs from /mv: mv validates completeness; siycftr fills gaps