A flat fan-shaped milled bed with five diverging grooved runners ending at a cube, a cone, a ribbed cylinder, a stepped pyramid and a twisted prism, a selector arm resting over the twisted prism; the diagonal cut ghosts the outermost runner in bone-grey line, and an indigo channel runs from the inlet through the splitter to the twisted prism.
Development

BeCreative

Single-pass divergent ideation via Verbalized Sampling: surfaces the strongest of several diverse candidates, or expands a seed corpus into an eval dataset

07
Workflows
01
Tool
02
References
15
Triggers

The Problem

Ask an AI to be creative and it defaults to the most likely output — the weighted average of everything it was trained on. That's the typicality bias baked into standard sampling. You get competent, safe, forgettable results: the blog angle everyone else would take, the product name that sounds like every other product name, the solution that looks like the obvious solution. The model isn't constrained by ideas; it's constrained by probability.

How This Skill Approaches It

Apply Verbalized Sampling (Zhang et al., 2024) combined with extended thinking. The model generates five internally diverse candidates with probability below 0.10 each — low-probability options it would normally never surface — then evaluates them and returns the strongest one. Seven workflows cover the range from quick brainstorming (StandardCreativity, IdeaGeneration) to boundary-pushing (MaximumCreativity), structured multi-path exploration (TreeOfThoughts), and domain-specific work. TechnicalCreativityGemini3 routes engineering problems to Gemini 3 Pro. SyntheticDataExpansion runs VS-Multi for growing a small seed corpus into a diverse N-example dataset for evals or training — useful when you have 5-20 examples and need 50-200 with genuine variety. The research shows 1.6-2.1x diversity increase and 25.7% quality improvement over standard sampling under controlled conditions.

Not for evolving ideas over several rounds, multi-cycle ideation, breakthrough ideas, ideate or id8 (HillClimb Ideate), new causal explanations of a phenomenon (Novelty), or adversarial critique (RedTeam)

In Action

What you say to your DA, and what the BeCreative skill actually does.

  • You say "be creative — i need a name for this security tool"
    Runs MaximumCreativity: generates five internally diverse naming candidates (p<0.10 each) via Verbalized Sampling during extended thinking, then returns the strongest option with brief reasoning.
  • You say "i have 10 eval examples, expand them to 50 with real variety"
    Runs SyntheticDataExpansion (VS-Multi): takes the seed corpus, generates batches via low-probability sampling across the distribution, and returns a diverse N-example dataset that doesn't cluster around the originals.

Inside the Skill

The thinking, frameworks, and architecture that distinguish this skill from a generic version of the same task.

Ideal State

  • Each run surfaces several genuinely distinct candidates, compared against each other rather than by self-reported probability, and names the strongest for the brief.
  • The candidates differ in kind (different assumptions, framings, formats or domains), never one idea in five outfits.
  • When diversity is load-bearing, candidates are generated or judged across a second vendor, the lane chosen by whether the brief is taste-judged (Anthropic) or correctness-judged (OpenAI).
  • SyntheticDataExpansion returns a corpus measurably more diverse than its seed set: a higher diversity index from Tools/Diversity.ts index than the seed.
  • A request to evolve ideas over several rounds reaches HillClimb Ideate, never a single pass here.

Intent Routing

Intent Trigger File
StandardCreativity "be creative", "think creatively", default creative tasks Intents/StandardCreativity.md
MaximumCreativity "maximum creativity", "most creative", "radically different" Intents/MaximumCreativity.md
IdeaGeneration "brainstorm", "ideas for", "solve this problem" Intents/IdeaGeneration.md
TreeOfThoughts "complex problem", "multi-factor", "explore paths" Intents/TreeOfThoughts.md
DomainSpecific "artistic", "business innovation", domain-specific Intents/DomainSpecific.md
TechnicalCreativity "technical creativity", algorithm or architecture ideas Intents/TechnicalCreativity.md
SyntheticDataExpansion "expand corpus", "synthetic data", "generate diverse examples", "create test set from these" Intents/SyntheticDataExpansion.md

"ideate", "id8", "evolve ideas", "breakthrough ideas", "multi-cycle ideation" and any request for several rounds route to HillClimb Ideate (skills/HillClimb/Intents/Ideate.md).

File Read when
ResearchFoundation.md the Verbalized Sampling evidence base behind the single-pass intents
Setup.md first-time setup: the OpenAI login TechnicalCreativity's second vendor needs, and a read-only check

Data & Config

User customizations, when present, live at <configRoot>/LIFEOS/USER/CUSTOMIZATIONS/SKILLS/BeCreative/ (configRoot is $CLAUDE_CONFIG_DIR when set, else ~/.claude). Any PREFERENCES.md, configurations, or resources there override the defaults; when the directory does not exist, proceed with skill defaults.

First-time setup for the TechnicalCreativity vendor split is in Setup.md.

Gotchas

  • Verbalized Sampling adds nothing on tasks with one correct answer. Use it where many answers are valid. A hill climb has one metric but many valid changes, so VS applies to its proposals (HillClimb Ideate), never to its keep decision (2026-10-11 analysis).
  • The model's self-reported probabilities are not evidence of novelty. Compare the candidates to each other.
  • A single model's "diverse" set shares that model's defaults. When diversity really matters (TechnicalCreativity, and HillClimb Ideate's rosters), generate or judge on a second vendor through LIFEOS/TOOLS/Inference.ts --vendor.
  • Taste-judged briefs (voice, naming, visual concepts) belong on the Anthropic lanes; correctness-judged ones (algorithms, architectures) on the OpenAI lanes. Pick the --vendor accordingly when an intent calls a model directly.

Examples

User: "Brainstorm products we could build from a patent portfolio"
→ IdeaGeneration
→ The most breakthrough of five approaches that rest on different assumptions, with why it beats the conventional answer
User: "Divergent scene concepts for a blog header illustration"
→ DomainSpecific
→ A scene concept that breaks a real editorial-illustration convention, chosen on that domain's criteria and justified against them
User: "Evolve ideas for a new featured content collection over a few rounds"
→ HillClimb Ideate (not this skill)
→ Generations from a code-drawn roster across bands and vendors, judge-scored with spread since there is no metric, the stopping condition named, each generation persisted

Workflows & Routing · 7

Each workflow is one job the skill runs. The trigger phrases route your request to the right one — this is the skill's routing table.

  1. 01
    StandardCreativity Intents/StandardCreativity.md

    be creative, think creatively, default creative tasks

  2. 02
    MaximumCreativity Intents/MaximumCreativity.md

    maximum creativity, most creative, radically different

  3. 03
    IdeaGeneration Intents/IdeaGeneration.md

    brainstorm, ideas for, solve this problem

  4. 04
    TreeOfThoughts Intents/TreeOfThoughts.md

    complex problem, multi-factor, explore paths

  5. 05
    DomainSpecific Intents/DomainSpecific.md

    artistic, business innovation, domain-specific

  6. 06
    TechnicalCreativity Intents/TechnicalCreativity.md

    technical creativity, algorithm or architecture ideas

  7. 07
    SyntheticDataExpansion Intents/SyntheticDataExpansion.md

    expand corpus, synthetic data, generate diverse examples, create test set from these

Tools · 1

Deterministic executables the workflows call — the code that does the real work, not prompt scaffolding.

  • Diversity.ts

How to Invoke

Say any of these to your DA and LifeOS activates the BeCreative skill automatically:

  • "be creative"
  • "brainstorm"
  • "divergent ideas"
  • "creative solutions"
  • "maximum creativity"
  • "tree of thoughts"
  • "radically different"
  • "name this"
  • "creative angle"
  • "technical creativity"
  • "novel ideas"
  • "expand this corpus"
  • "synthetic data"
  • "generate diverse examples"

Or invoke explicitly:

Skill("BeCreative")

References · 2

Auxiliary files the skill loads at runtime — frameworks, guides, configs.

  • ResearchFoundation
  • Setup

References & Credits

The thinkers, books, frameworks, and research this skill is built on. The ideas belong to them — the integration belongs to LifeOS.

Want LifeOS to do this for you?

Install LifeOS on your machine — your DA gets the BeCreative skill plus 96 others, all hooked into one Life OS.