Pivot Idea

Diagnose why an idea scored badly and generate evidence-backed pivots with projected scores

Latifox 8d0b6d5 1.2 KB Updated

File contents

Run the workflow in workflows/pivot-optimization.md.

Arguments: $ARGUMENTS

If the user named a single variable to change (pricing, audience, channel, platform, feature emphasis) - for example "what if I charged per client instead of per seat" - run the Micro-pivot path: skip weakness detection, generate exactly one option with pivot_scope: micro and variables_changed: 1, then re-score it. If that named change fails the Minimum Viable Pivot Criteria, say so plainly rather than substituting a different pivot.

Otherwise run the standard chain: confirm the current score, detect weaknesses and their root causes, generate two or three options, and re-score the best one.

Apply the Slug Rule when deciding where output lands. One variable changed updates the idea in place. Two variables, a change of business_model, or a change of the core problem creates memory/ideas/<slug>-<pivot-word>/ with pivot_of set, and pauses the original with superseded_by.

Only addressable_weaknesses are pivot targets. A knowledge gap means a dimension file is missing or thin; send it back to that research skill instead.

Latifox/find-me-saas/tree/main/.claude/skills/pivot-idea commit 8d0b6d55b2

Frequently asked questions

npx skillmds@latest add latifox/pivot-idea