SYCS - So You Can See

Input: $ARGUMENTS

Core Principles

Implications branch exponentially. Every statement implies multiple things, each of which implies multiple things. Without principled pruning, you drown in implications. The skill must bound the expansion while preserving the important branches.
Depth degrades reliability. First-order implications are usually solid. Second-order implications are plausible. Third-order and beyond are speculative. Track the depth and mark the reliability gradient.
Not all implications are logical. Some implications are logical necessities (if A then necessarily B). Others are empirical predictions (if A then probably B). Others are social/practical consequences (if A then people will likely do C). Each type has different confidence.
The most important implications are non-obvious. The obvious implication of “we’re cutting the budget” is “we’ll spend less.” The non-obvious implication might be “we’ll lose our best engineer who joined for the project budget enables.” Surface the non-obvious ones.
Testability makes implications useful. An implication you can test is actionable. An implication you can’t test is speculation. Flag which is which.

Phase 1: Source Statement

[Y1] SOURCE: [the statement to expand, quoted]
[Y2] CONTEXT: [relevant context that affects implications]
[Y3] STATEMENT_TYPE: [claim | decision | observation | prediction | policy]

Phase 2: First-Order Implications

Direct implications — what follows immediately from the source statement.

[Y4] IMPLICATION: [what follows] — TYPE: [logical | empirical | practical | social]
  CONFIDENCE: [high | medium | low]
  OBVIOUS: [yes | no]
  TESTABLE: [yes — how | no]
[Y5] IMPLICATION: [what follows] — TYPE: [type]
  CONFIDENCE: [level]
  OBVIOUS: [yes/no]
  TESTABLE: [yes/no]
...

Implication Types

Type	What It Means	Confidence Basis
Logical	Necessarily follows from the statement	Deductive — if A then must B
Empirical	Usually follows based on evidence/experience	Inductive — if A then probably B
Practical	Follows as a consequence of action/policy	Consequential — if A then action B needed
Social	Follows from how people typically respond	Behavioral — if A then people will likely B

Phase 3: Second-Order Implications

For each high-confidence or high-impact first-order implication, expand one level deeper:

[Y-N] FROM [Y-ref]: [second-order implication]
  TYPE: [logical | empirical | practical | social]
  CONFIDENCE: [typically lower than parent]
  IMPACT: [high | medium | low]
  TESTABLE: [yes — how | no]

Pruning Rules

Expand a first-order implication to second order ONLY if:

Its confidence is medium or higher, AND
Its impact is medium or higher, OR
It’s non-obvious (surprising implications deserve expansion even if low impact)

Do NOT expand if:

First-order confidence is low (speculating on speculation)
The implication is obvious and low-impact (not worth the depth)

Phase 4: Impact Flagging

From all implications (first and second order), identify the highest-impact ones:

[Y-N] HIGH-IMPACT IMPLICATIONS:
  1. [Y-ref]: [implication] — IMPACT: [why this matters most]
     TESTABLE: [yes — test | no]
     ACTION_IF_TRUE: [what to do about it]
  2. [Y-ref]: [implication] — IMPACT: [why]
     TESTABLE: [yes/no]
     ACTION_IF_TRUE: [action]
  3. [Y-ref]: [implication] — IMPACT: [why]
     ...

Non-Obvious Flag

[Y-N] NON-OBVIOUS IMPLICATIONS:
  1. [Y-ref]: [implication] — WHY_NON_OBVIOUS: [why most people would miss this]
  2. [Y-ref]: [implication] — WHY_NON_OBVIOUS: [why]

Phase 5: Output

IMPLICATION EXPANSION
=====================

SOURCE: [quoted statement]
CONTEXT: [relevant context]

FIRST-ORDER (direct implications):
  [Y4] [implication] — [type] — Confidence: [level] — Testable: [yes/no]
  [Y5] [implication] — [type] — Confidence: [level] — Testable: [yes/no]
  ...

SECOND-ORDER (implications of implications):
  From [Y4]:
    [Y-N] [implication] — [type] — Confidence: [level]
  From [Y5]:
    [Y-N] [implication] — [type] — Confidence: [level]
  ...

HIGH-IMPACT (prioritized):
  1. [implication] — [why high impact] — Test: [how]
  2. [implication] — [why high impact] — Test: [how]

NON-OBVIOUS:
  1. [implication] — [why most people miss this]

READY FOR:
- /ht [implication] — to formulate a testable hypothesis
- /ar [implication] — to explore what follows if this implication is right
- /aw [implication] — to stress-test a high-impact implication
- /claim [implication] — to verify a specific implication

Failure Modes

Failure	Signal	Fix
Unbounded expansion	30+ implications listed with no pruning	Apply pruning rules at each depth level
Only obvious implications	Every implication is “well, duh”	Specifically search for non-obvious consequences
Confidence not tracked	All implications treated as equally likely	Mark confidence at each depth level
Depth-blind	Third-order implications treated with same confidence as first-order	Confidence typically decreases with depth
Type conflation	Logical necessity confused with empirical prediction	Apply type definitions — does B NECESSARILY follow from A?
No testability check	Implications listed without actionability	Flag testable vs not-testable for each
Impact not prioritized	Flat list with no high-impact flagging	Always identify the top 3 highest-impact implications

Depth Scaling

Depth	Min First-Order	Max Depth	Min High-Impact	Min Non-Obvious
1x	3	1st order only	1	0
2x	5	2nd order	2	1
4x	8	3rd order	4	2
8x	12	4th order	6	4

Default: 2x. These are floors.

Pre-Completion Checklist

Source statement clearly identified with context
First-order implications extracted with type and confidence
Pruning rules applied before second-order expansion
Second-order implications have lower confidence than parents
High-impact implications flagged with reasoning
Non-obvious implications specifically sought and identified
Testability assessed for each implication
Action recommendations for testable high-impact implications

Integration

Use from: any statement analysis, strategic planning
Routes to: /ht, /ar, /aw, /claim for testing specific implications
Complementary: /aex (extracts hidden assumptions — different from implications)
Differs from /aex: aex finds what must be TRUE for a claim to hold; sycs finds what FOLLOWS from a claim
Differs from /ar: ar explores what follows if something is RIGHT; sycs expands implications neutrally

sycs - So You Can See