Lateral thinking
Structured thinking skills for AI agents: brainstorming, decision support, systems thinking, and strategy. Built for Claude Code & Cursor.
npx -y skills add ysskrishna/ai-agent-skills --skill lateral-thinkingAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 3 stars3 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
What its author says it does
Copied from the file, not written here
Use this skill when the user asks for lateral thinking (including naming it or directing use/apply/run with obvious misspellings; decisive) or wants to break a sticky pattern—provocations, concept fans, challenge questions, and deliberate reframes to escape local optima. Use when ideation feels stuck, options feel incremental, or they want surprising angles (including after conventional brainstorming stalls), even if they never say "lateral thinking" or "provocation". Skip when they need a single compliant rule lookup, tight logical proof with no exploration phase, or no appetite for playful what-if moves.
The file declares its own license as MIT. That is the author’s claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.
SKILL.md
3.3 KB, as published. Nobody here has run it
Lateral Thinking
Movement before judgment. Treat provocative lines as stepping stones, not final positions.
How to run it with this skill: one clearly headed section per move in the default order from Setup (Provocation (Po) → Extract principle → Concept fan → Candidates), or the alternate path the user chose.
Setup (run before starting)
In one short block:
- Stuck point — what pattern or assumption feels rigid?
- Default pass — Provocation (Po) → Extract principle → Concept fan → Candidates (state this line)
If safety or legal constraints are unclear, ask at most 3 questions in one message before generating provocations. Note any remaining gaps in plain language (no bracket tags in Setup).
If the user asks for random stimulus instead of a Po, replace the Provocation step with Random entry: one random noun or image, then several bridges (property → analogy → mechanism → application) into ideas.
If the user asks to challenge implicit rules, replace Provocation / Extract principle / Concept fan with: list implicit rules → for each, Challenge / Reframe / Test (smallest experiment) → Escape (two ways to sidestep the load-bearing rule) → then Candidates only from that thread.
The Moves
Provocation
Offer one primary Po (statement can be illogical or impossible) formatted:
Po: … — Movement idea: what useful idea does this suggest if we treat it as a springboard?
Extract principle
From the movement idea, state one operating principle that could work in the real world (even if the Po itself is absurd).
Concept fan
Widen from the principle along three directions — broaden, deepen, redirect — with several branches per direction unless the user wants more.
Candidates
Several actionable ideas. Tag each: near-term / stretch / experimental.
Execution Rules
- Never present a Po as a factual recommendation; always pair with Movement or Principle.
- If a provocation touches safety, ethics, or law, stop and surface the conflict explicitly — do not rationalize harm.
- One response for all moves unless interactive pacing is requested.
Checklist (verify before responding)
- Setup: stuck point + default pass (note if Random entry or rule-challenge path used)
- Po paired with movement; or alternate path completed as requested
- Concept fan has three directions with multiple branches each
- Candidates tagged near-term / stretch / experimental
- No Po presented as literal policy