YCSHIKFMIF - You Can See How I Keep Finding More It’s Fun

Input: $ARGUMENTS

Core Principles

1. Expansion loops are valuable AND dangerous. The excitement of finding more ideas is a genuine signal that the space is rich. But unbounded expansion delays execution indefinitely. Both things are true simultaneously.

2. Batch, don’t stop. The answer to “I keep finding more” isn’t “stop finding” — it’s “batch what you have, explore the next batch later.” This preserves the expansion energy while creating execution boundaries.

3. Not all new ideas are equal. Some expand into rich territory. Others are dead ends wearing exciting costumes. The batch must be sorted by promise before exploration.

4. The fun is real information. Excitement about finding more ideas signals a rich possibility space. Don’t dismiss it. But channel it — fun without structure becomes procrastination.

5. Stop conditions prevent infinite exploration. Every batch needs a “done exploring this batch” condition. Without it, each batch expands into its own unbounded loop.

Phase 1: Expansion Capture

Capture the current state of the expansion loop:

[K1] EXPANSION_CONTEXT: [what the user was originally trying to do]
[K2] IDEAS_FOUND_SO_FAR: [list all ideas currently on the table]
[K3] IDEA_COUNT: [N ideas]
[K4] EXPANSION_RATE: [accelerating | steady | decelerating]
[K5] EXPANSION_SOURCE: [where new ideas are coming from — brainstorming, research, conversation, association]
[K6] TIME_IN_EXPANSION: [how long has the user been finding more?]

Phase 2: Idea Assessment

For each idea, quick-score:

[K-N] IDEA: [description]
  PROMISE: [high | medium | low] — [why]
  NOVELTY: [genuinely new | variant of existing | redundant]
  EFFORT_TO_EXPLORE: [small | medium | large]
  DEPENDENCY: [standalone | depends on [other idea]]

Quick Filters

Phase 3: Batching

Sort ideas into batches:

[K-N] BATCH 1 — EXPLORE NOW (current session):
  Size: [N ideas]
  Stop condition: [when this batch is "done"]
  Ideas:
    1. [idea] — PROMISE: [level] — EFFORT: [level]
    2. [idea] — PROMISE: [level] — EFFORT: [level]
    ...

[K-N] BATCH 2 — EXPLORE NEXT (deferred, not abandoned):
  Size: [N ideas]
  Trigger: [when to open this batch]
  Ideas:
    1. [idea] — PROMISE: [level]
    2. [idea] — PROMISE: [level]
    ...

[K-N] PARKED (interesting but not now):
  1. [idea] — WHY_PARKED: [reason]
  2. [idea] — WHY_PARKED: [reason]

[K-N] DROPPED (not worth pursuing):
  1. [idea] — WHY_DROPPED: [redundant | low promise | too expensive]

Batch Size Rules

Phase 4: Stop Conditions

[K-N] BATCH 1 STOP CONDITIONS:
  TIME_LIMIT: [stop after N minutes/hours]
  IDEA_LIMIT: [stop after exploring N ideas in this batch]
  DIMINISHING_RETURNS: [stop when new findings per idea drop below threshold]
  SUFFICIENT_COVERAGE: [stop when N% of the original question is covered]
  EXECUTION_READY: [stop when you have enough to act on]

[K-N] EXPANSION_LOOP_DETECTOR:
  IF during Batch 1 exploration, new ideas spawn at rate > [threshold]:
    → PAUSE exploration
    → Capture new ideas to Batch 2
    → Return to Batch 1 stop condition check
    → DO NOT expand Batch 1

Phase 5: Output

EXPANSION MANAGEMENT
====================

ORIGINAL GOAL: [what the user was trying to do]
IDEAS FOUND: [N total]
EXPANSION RATE: [level]

BATCH 1 — EXPLORE NOW ([N] ideas):
  1. [idea] — Promise: [level] — Effort: [level]
  2. [idea] — Promise: [level] — Effort: [level]
  ...
  STOP WHEN: [condition]

BATCH 2 — EXPLORE NEXT ([N] ideas):
  1. [idea] — Promise: [level]
  ...
  OPEN WHEN: [trigger]

PARKED ([N] ideas): [listed for reference, not abandoned]
DROPPED ([N] ideas): [with reasons]

LOOP PROTECTION: If new ideas emerge during Batch 1, capture to Batch 2.
Do NOT expand Batch 1.

READY FOR:
- Execute Batch 1 ideas
- /iagca — if Batch 1 is still too large
- /awtlytrn — to estimate limits for batch execution
- /ro — to reorder ideas within a batch

Failure Modes

Depth Scaling

Default: 2x. These are floors.

Pre-Completion Checklist

• All current ideas captured (not just some)
• Each idea assessed for promise, novelty, and effort
• Ideas sorted into batches with size limits
• Batch 1 has explicit stop conditions
• Batch 2 has explicit trigger conditions
• Parked ideas preserved (not deleted)
• Dropped ideas have reasons
• Loop detector specified for re-expansion during Batch 1
• Original goal is still visible (not lost in expansion)

Integration

• Use from: brainstorming sessions, research, any ideation that’s expanding
• Complementary: /iagca (when scope needs aggressive compression)
• Complementary: /awtlytrn (to estimate execution limits for batch)
• Differs from /iagca: iagca compresses scope urgently; ycshikfmif structures ongoing expansion
• Differs from /ma: ma generates ideas deliberately; ycshikfmif manages organic expansion