lateral-thinking skill
Ten skills for Claude Code and Codex: a router diagnoses a stuck idea and runs the matching lateral thinking technique
- 330 ★ on the repository
- Last commit
- MIT
Minimal access
Why this level
- The set consists of Markdown instructions, runs nothing and does not touch the network
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/lateral-thinkingDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewMinimal accessThe tool runs nothing and makes no network calls.
- The set consists of Markdown instructions, runs nothing and does not touch the network
This is third-party code. Review the repository files before installing.
What it does
The lateral-thinking set of ten skills gives the agent eight structured techniques for finding a new product direction. A vanilla brainstorm from an agent returns three sensible features that turn out to be one idea in different wrappers. The `lateral` skill acts as a router: it asks what you are stuck on, picks exactly one technique and runs it from the instructions in a sibling folder. The techniques come from known methods. Random stimulus, provocation and concept fan come from Edward de Bono, and so do the six thinking hats. SCAMPER was devised by Bob Eberle. Forced analogy comes from the Synectics school, and assumption inversion from classical dialectic. Every skill carries an abandonment rule: a condition under which the agent drops an idea and shows you that it did. After a batch it looks for a shared meta-pattern. One more skill, `post`, turns a topic or someone else's post into draft publications.
Who it is for. For product managers, founders and developers who ask an agent for ideas and get predictable answers.
Good fit when
- Users create one project and never return, and the question of what brings them back yields only email reminders
- Every brainstorm idea looks like the last one, which suits `random-stimulus` with random stimuli from unrelated domains
- A rule or constraint feels unbreakable: `provocation` states deliberately wrong assertions and extracts movement from them
- A decision is being made too fast and everyone agrees: `six-hats` runs it through six separate passes from facts to synthesis
- You have one idea and need variants: `scamper` checks it against seven operations and shows empty results as empty
Not a fit when
- You need to debug code, review a pull request or build a feature: the router does not take such tasks
- You want one quick answer without a multi-step session: techniques unfold into a dialogue and need your input
- You installed only the `lateral` skill without sibling folders: it will run a condensed technique without stimulus pools and worked examples
- You need a finished business strategy with calculations: the set widens options and does not size a market
Example request
Our users create one project and never return. Use the `lateral` skill and find a product direction that makes the first project worth coming back to.Limitations
- The agent decides on its own whether to call the skill, and different agents behave differently. It is more reliable to name the skill directly: use `lateral`.
- The package must be installed whole, otherwise the router runs in a reduced mode.
- Six Thinking Hats is a registered trademark of the de Bono Group, and the implementation here is an independent educational one.
- The set gives structured divergence with quality rules, and it does not guarantee a good idea.
How to disable. Delete the skill folders from your agent's skills directory. If you installed the plugin in Claude Code, run `/plugin uninstall lateral-thinking@lateral-thinking`.
FAQ
How do I install lateral-thinking in Claude Code?
There are two routes. Universal: `npx skills add danium/lateral-thinking`. Through the plugin: `/plugin marketplace add danium/lateral-thinking`, then `/plugin install lateral-thinking@lateral-thinking`. For Codex, ask `$skill-installer` to install the skill from the repository address.
Which technique should I pick when all ideas look alike?
The router table lists `random-stimulus` for that symptom. If you suspect the wrong problem is being solved, `concept-fan` fits, and for a derivative but working idea, `analogy`.
What is an abandonment rule?
It is a condition under which the agent throws an idea away and shows the abandonment. The authors argue that visible failures make the other results more credible. An agent whose seven ideas are all strong is more likely copying the shape of a good answer.
Can I use the skills for code analysis?
No. Debugging, review and implementation are not creative targets, and the router declines. A request to reinvent the process itself is fine, for example a code review ritual.
What does the `post` skill do?
It takes a topic or someone else's post, routes it through the set of techniques and shapes the result into drafts you can send. It is an applied skill that sits next to the eight techniques.
Discussion
No questions or reviews yet.
Related
SkillsA catalog of 28 mental-model skills for agents: diagnosis, decisions, systems, risk, strategy and idea generation
An analytical skill that tests whether users will find value in a product idea: scenario, conditions, timing and delivery, with solidity ratings
Lenny's Product Skills
Lenny's Product Skills for Claude Code
A set of product skills for Claude Code and SKILL.md agents: strategy, discovery, launch, growth and team work
Second Brain Skills
Claude Code Second Brain Skills
A set of Claude Code skills for knowledge work: brand and voice, an MCP client, PPTX slides, Remotion video, SOPs and a skill creator
A set of 13 SKILL.md skills: research, a product doc, decks, articles and audits for any compatible agent
A large library of product frameworks as skills: discovery, strategy, PRDs, prioritization and unit economics
Comparison with similar tools
Entries that share the main search query, and how they differ.
| Entry | Official | Repository stars | Main difference |
|---|---|---|---|
| lateral-thinking (this entry) | no | 330 | |
| ADHD | no | 4,484 | parallel divergent ideation under cognitive frames with pruning |
| Compound Engineering | no | 25,769 | a brainstorm, plan, build, review and capture loop for engineering work |
README in short
A demo compares answers to one question about user retention without a skill and with the `random-stimulus` technique. The author walks through a live `scamper` session and lists the method sources.
README badge
Original SKILL.md
--- name: lateral description: Lateral thinking toolkit router — when you're stuck, going in circles, need fresh ideas, or standard brainstorming keeps producing predictable results, this diagnoses the symptom and applies the right technique from the toolkit (random-stimulus, provocation, inversion, concept-fan, analogy, scamper, six-hats, worst-idea). Triggers include "lateral thinking", "I'm stuck on a creative problem", "we're going in circles", "need fresh ideas", "try a different angle", "break out of the box", "ideas all feel the same", "predictable brainstorm". Do NOT use for analytical work like debugging, code review, or implementation tasks. --- # Lateral ## What this does Being stuck has shapes. Ideas that all feel the same is a different problem from a constraint that feels unbreakable, which is different again from suspecting you are solving the wrong problem. Each shape has a technique that fits it. This skill diagnoses the symptom, picks exactly one technique, and runs it inline. It is a router, not a technique of its own. ## Decision table | Symptom | Technique | |---|---| | Ideas all feel the same; brainstorm output is predictable | `random-stimulus` | | A constraint or rule feels unbreakable | `provocation` | | Requirements assume things nobody has questioned | `inversion` | | We might be solving the wrong problem | `concept-fan` | | The solution works but feels derivative | `analogy` | | We have one idea and need variations | `scamper` | | A decision is being made too fast / everyone agrees | `six-hats` | | Everything feels timid, safe, cautious | `worst-idea` | ## Routing procedure 1. **Diagnose.** Match the user's symptom to the table. If the symptom is unclear, ask exactly one focused question — do not interrogate. 2. **Pick exactly one technique.** Never route to a second technique in the same pass. If another looks promising, offer it as a next move once the first has finished.