Skip to content

TheGoat395/Codex-Skills / skills/creativity

creativity

Verified

Generate original ideas or escape fixation.

references/creative-search-methods.md

View on GitHub
# Creative Search Methods

Use this reference for substantial ideation, a stale direction, or a problem where the first answers are predictably generic.

## Generate preinventive structures

Before writing complete solutions, create incomplete material that can be interpreted in several ways:

- spatial or process arrangements;
- category combinations;
- visual or verbal patterns;
- functions detached from their normal object;
- mechanisms from another field;
- constraints, contradictions, and impossible wishes;
- partial systems, sequences, and transformations.

Then explore what each fragment could become. This generation-exploration cycle follows the Geneplore model: generative processes create preinventive structures and exploratory processes interpret and elaborate them under task constraints.

## Flexibility operators

Use these to reach a different region of the search space:

| Operator | Move |
|---|---|
| Reframe | Change the problem, beneficiary, unit of value, or definition of success |
| Invert | Reverse the sequence, control, visibility, ownership, or default assumption |
| Translate | Move a mechanism across domains while discarding surface features |
| Change representation | Convert text to space, process to object, data to narrative, or instruction to interaction |
| Change scale | Design for one, many, instant, long-term, microscopic, or system-wide use |
| Change actor | Let a different stakeholder initiate, decide, verify, or benefit |
| Remove the center | Eliminate the dominant component and rebuild the value through another path |
| Add a productive constraint | Force one material, gesture, channel, rule, or proof form to organize the concept |
| Counterfactual | Remove the assumed tool, budget, convention, or environment |

After each operator, identify the changed mechanism. Reject output that only changes vocabulary.

## Persistence operators

Use these after a branch appears promising but obvious:

1. list the first accessible answers and mark them as the surface layer;
2. choose one mechanism, not the whole concept;
3. generate consequences of that mechanism at three levels of depth;
4. identify the weakest component and produce several replacements;
5. vary one structural parameter while holding the rest constant;
6. trace second-order effects, edge cases, and unexpected uses;
7. develop the least obvious viable variant, not merely the favorite;
8. stop when variants repeat or a prototype becomes more informative.

Persistence is not polishing. It explores a chosen conceptual region more deeply.

## Analogy transfer

Transfer relations, not appearances:

```text
Source domain:
Useful mechanism:
Why the mechanism works there:
Structural match in the target:
Required adaptation:
Surface features to discard:
Where the analogy breaks:
```

A distant analogy is valuable when the mechanism transfers and the target changes meaningfully. Novel vocabulary around an unchanged default is not transfer.

## Combination matrix

When fragments may interact, cross two independent dimensions, such as:

- mechanism x audience role;
- proof form x delivery channel;
- constraint x interaction;
- emotional tension x information structure;
- physical metaphor x system behavior.

Inspect emergent combinations. Do not mechanically fill every cell. Develop only combinations with a coherent reason to exist.

## Representation switching

When stuck, express the problem as a different object:

- causal diagram;
- journey or sequence;
- opposing forces;
- resource flow;
- set of transformations;
- game with actors, incentives, and rules;
- physical space;
- conversation;
- reversible experiment.

The new representation should expose relationships hidden by the old one. Switch back before delivery so the final output fits the real medium.

## Fixation controls

Design fixation research shows that examples can cause designers to reproduce problematic features. Reduce that risk by:

- generating functional requirements before viewing examples;
- inspecting several diverse examples instead of one canonical answer;
- abstracting each example to a mechanism;
- listing forbidden surface transfers;
- creating a non-example or counterexample;
- testing whether a candidate can be derived from the problem without the exemplar;
- comparing candidates for structural similarity, not just visual similarity.

References remain useful evidence and inspiration. The goal is controlled transfer, not reference avoidance.

## Search ledger

For a consequential creative problem, use a compact ledger:

```text
Search objective:
Fixed constraints:
Baseline to escape:

Branch | Mechanism | What is genuinely different | Promise | Main constraint | Next test | Status
```

Keep the ledger small. It should prevent repeated ideas and forgotten branches, not become a project-management artifact.

## Research basis

- Ronald Finke, Thomas Ward, and Steven Smith, *Creative Cognition* (1992): the Geneplore model separates generation of preinventive structures from their exploration and interpretation.
- Carsten de Dreu, Matthijs Baas, and Bernard Nijstad, *Hedonic Tone and Activation Level in the Mood-Creativity Link* (2008): creativity in their studies followed distinguishable flexibility and persistence pathways.
- David Jansson and Steven Smith, *Design Fixation* (1991): exposure to example designs increased reproduction of example features, including flawed ones, in the reported design tasks.
- Steven Smith, Thomas Ward, and Jay Schumacher, *Constraining Effects of Examples in a Creative Generation Task* (1993): examples can constrain generated ideas toward demonstrated properties.
slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack."Stars: 198kUpdated Claude CodeCodexVerifiedInstall
ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read.Stars: 194kUpdated Claude CodeCodexVerifiedInstall
ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report.Stars: 194kUpdated Claude CodeCodexVerifiedInstall
ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt.Stars: 194kUpdated Claude CodeCodexVerifiedInstall