generative-thinking

Break out of a locked problem frame by picking one disciplined generative move — reframe, provocation (Po), random stimulus, SCAMPER, inversion, structured analogy, constraint play, or family traversal — and committing to it before evaluating.

The skill is the interrupt for fixation, not a brainstorming workshop. Pick one move matched to the diagnosis, run it, over-generate and tail-sift (write next to each candidate the probability the old frame would have reached it; drop the likely ones), apply the fire test (could this output have been produced without the move?), stop.

See SKILL.md for the diagnostic table, per-technique recipes with success signals, LLM-specific sourcing notes, and canonical references. references/stimulus-vocabulary.md is a 128-noun deck for the hashed random stimulus.

Provenance

• Lateral thinking, Po, random stimulus — Edward de Bono, Lateral Thinking (1970), Serious Creativity (1992)
• SCAMPER — Bob Eberle (1971), built on Alex Osborn's Applied Imagination (1953)
• Oblique Strategies — Brian Eno & Peter Schmidt (1974–2001)
• Functional fixedness (the bias this skill counters) — Karl Duncker (1935/1945); candle problem
• LLM-specific design fixation — Wadinambiarachchi et al., CHI 2024 (doi:10.1145/3613904.3642919): AI-generated examples during ideation increase fixation and reduce fluency, variety, originality vs. no-support baseline. Motivates the LLM-sourcing section.
• Family traversal — Balestriero, Pesenti & LeCun, arXiv:2110.09485 (interior ≠ novelty; limits extrapolate, mixtures average)
• Tail sift (typicality bias, verbalized-probability prompting) — Jiayi Zhang, Simon Yu, Derek Chong, Anthony Sicilia, Michael R. Tomz, Christopher D. Manning & Weiyan Shi, Verbalized Sampling, arXiv:2510.01171
• Hashed internal random stimulus — Kou Misaki & Takuya Akiba (Sakana AI), String Seed of Thought, arXiv:2510.21150
• Structured analogy — Andrew Shen, Shaul Druckmann & James Zou (Stanford), Unlocking LLM Creativity in Science through Analogical Reasoning, arXiv:2605.11258, built on Dedre Gentner's structure-mapping theory (Cognitive Science, 1983)

The three 2025–2026 papers were pointed out by Stuart Gray; 0.3.0 is the result.

Complements

• challenging — evaluates after generation
• convening-experts — synthesizes multiple role-based viewpoints
• tiling-tree — exhaustive MECE partitioning of a solution space

This skill generates distance, not judgment, coverage, or consensus.

Skill folder: generative-thinking

Release of generative-thinking version 0.3.0

📥 Download & Install

⬇️ Download generative-thinking.zip

To install:

1. Click the download link above (ignore the "Source code" archives below - they're auto-generated by GitHub)
2. Go to Claude.ai Skills Settings
3. Upload the downloaded ZIP file
4. Requires paid Claude Pro or Team account

See official documentation for more details.

Recent Changes

2b2ec48 feat(generative-thinking): 0.3.0 — tail sift, hashed random stimulus, structured analogy (#797)
ae85811 docs: Update CHANGELOG.md for released skills