IAGCA - I Am Getting Carried Away
Input: $ARGUMENTS
Core Principles
-
Sprawl feels like progress. Generating ideas, exploring tangents, and building elaborate plans all feel productive. They are NOT productive if they delay execution. The skill must distinguish genuine exploration from execution avoidance.
-
The highest-leverage action is rarely the most exciting one. Exciting ideas draw attention. But the thing that moves the project forward most is often boring, concrete, and small. Find it.
-
Defer, don’t delete. Loss aversion makes people resist cutting scope. The fix isn’t deleting ideas — it’s moving them to a “later” list. Everything of value is preserved. But only 2-3 things get done NOW.
-
Sprawl signals richness. Getting carried away means the space is interesting. That’s good information. But richness needs structure or it becomes chaos.
-
The scope compression is temporary. You’re not deciding what matters forever. You’re deciding what to do in the next 1-3 actions. Everything else waits.
Phase 1: Sprawl Assessment
[G1] ORIGINAL_GOAL: [what the user was originally trying to accomplish]
[G2] CURRENT_BRANCHES: [list every idea, tangent, and sub-project currently active]
[G3] BRANCH_COUNT: [N active branches]
[G4] SPRAWL_SEVERITY: [mild (5-8) | moderate (8-15) | severe (15+)]
[G5] SPRAWL_TYPE: [depth (going deeper on one thing) | breadth (spreading across many things) | both]
Phase 2: Branch Scoring
For each branch, score on two dimensions:
[G-N] BRANCH: [description]
NEAR_TERM_VALUE: [high | medium | low] — [what does this accomplish in the next 1-3 actions?]
EXECUTION_READINESS: [ready | needs-prep | blocked | unclear]
RELATIONSHIP_TO_GOAL: [direct | supporting | tangential | unrelated]
Scoring Matrix
| Near-Term Value | Execution Readiness | Action |
|---|---|---|
| High | Ready | DO NOW |
| High | Needs prep | PREP THEN DO |
| High | Blocked | UNBLOCK or DEFER |
| Medium | Ready | DO IF TIME |
| Medium | Not ready | DEFER |
| Low | Any | DEFER |
| Any | Unclear | CLARIFY or DEFER |
| Tangential/Unrelated | Any | DEFER |
Phase 3: Scope Compression
[G-N] DO NOW (maximum 3):
1. [branch] — WHY_NOW: [why this is the highest-leverage action]
2. [branch] — WHY_NOW: [reason]
3. [branch] — WHY_NOW: [reason]
[G-N] DEFERRED (preserved, not deleted):
1. [branch] — DEFERRED_BECAUSE: [reason] — REVISIT_WHEN: [condition]
2. [branch] — DEFERRED_BECAUSE: [reason] — REVISIT_WHEN: [condition]
...
[G-N] EXECUTION ORDER:
Step 1: [action] — [expected time]
Step 2: [action] — [expected time]
Step 3: [action] — [expected time]
Compression Rules
- Max 3 in DO NOW. If you can’t pick 3, the goal isn’t clear enough — route to
/gu. - Everything else is DEFERRED, not deleted. The user keeps every idea. Just not now.
- DO NOW items must be concrete actions, not “explore X” or “think about Y.”
- If no branch scores high: The sprawl might be a signal that the goal is wrong. Route to
/want.
Phase 4: Sprawl Source Check
Why did the user get carried away? Knowing the source prevents recurrence.
[G-N] SPRAWL_SOURCE: [one of the following]
| Source | Pattern | Prevention |
|---|---|---|
| Rich space | Domain has many genuine opportunities | Use /ycshikfmif to batch |
| Avoidance | Exploring is easier than executing | Name the avoided task; start with it |
| Unclear goal | Everything seems relevant because the goal is vague | Run /gu to clarify |
| Perfectionism | Need to cover everything before acting | Accept good enough; iterate later |
| External triggers | New inputs keep arriving (messages, ideas, requests) | Batch inputs; process on schedule |
Phase 5: Output
SCOPE COMPRESSION
=================
ORIGINAL GOAL: [restated]
BRANCHES ACTIVE: [N]
SPRAWL SEVERITY: [level]
SPRAWL SOURCE: [identified pattern]
DO NOW (max 3):
1. [concrete action] — [why this is highest leverage]
2. [concrete action] — [why this is next]
3. [concrete action] — [why this is third]
EXECUTION ORDER:
[action 1] → [action 2] → [action 3]
DEFERRED (preserved — not lost):
1. [idea] — Revisit when: [condition]
2. [idea] — Revisit when: [condition]
...
SPRAWL PREVENTION:
[what to do differently to avoid re-sprawl]
READY FOR:
- Execute the DO NOW list
- /ycshikfmif — if deferred ideas need batching
- /awtlytrn — if unsure about limits for DO NOW
- /gu — if sprawl source is unclear goal
Failure Modes
| Failure | Signal | Fix |
|---|---|---|
| DO NOW too large | More than 3 items | Cut to 3. If you can’t, the goal is unclear |
| All deferred | Nothing in DO NOW | If no branch is worth doing now, question the goal |
| Ideas deleted | User’s ideas removed instead of deferred | Everything goes to DEFERRED or DO NOW. Nothing is deleted |
| Vague actions | DO NOW contains “explore X” or “think about Y” | Actions must be concrete — “write the first draft of X” |
| Sprawl source ignored | Compression without understanding why it happened | Identify the source or compression will be temporary |
| Recurrence not addressed | Same sprawl pattern repeats next session | Prevention recommendation must address the root source |
Depth Scaling
| Depth | Min Branches Assessed | Scoring Dimensions | Sprawl Source | Prevention |
|---|---|---|---|---|
| 1x | 5 | Value only | Not checked | No |
| 2x | 10 | Value + readiness | Identified | Basic |
| 4x | 15 | Value + readiness + goal relationship | Identified + tested | With specific strategy |
| 8x | All | Full scoring matrix | Full analysis | With monitoring plan |
Default: 2x. These are floors.
Pre-Completion Checklist
- All active branches listed (not just some)
- Each branch scored on near-term value and execution readiness
- DO NOW contains maximum 3 concrete actions
- Everything else is DEFERRED (not deleted)
- Deferred items have revisit conditions
- Sprawl source identified
- Prevention strategy provided
- DO NOW items are concrete actions, not vague explorations
Integration
- Use from: any time the user recognizes they’re sprawling
- Complementary:
/ycshikfmif(structures ongoing expansion — less urgent) - Complementary:
/awtlytrn(estimates limits for the compressed scope) - Differs from
/ycshikfmif: ycshikfmif batches expansion; iagca compresses urgently - Differs from
/awtlytrn: awtlytrn estimates limits; iagca forces scope reduction - Routes to:
/guif sprawl source is unclear goal,/wantif no branch is worth doing