evolve-solution-population

SkillDev tools

Lets your agent improve solutions by evolving a pool of candidates, keeping diverse options instead of settling on one early.

Available today. Use it from your connected AI after setup.

Add ahel to your AI once: Claude, ChatGPT, Cursor, Claude Code or Codex. Then ask it to use this.

Then ask your AI: use the evolve-solution-population skill

About this skill

Maintain a diverse population of candidate solutions, mutate/recombine variants, select under explicit fitness and novelty criteria, and iterate while preserving useful niches rather than collapsing prematurely to one winner.

What this skill tells your AI

The instructions your AI receives, as published by yogsoth-ai/de-anthropocentric-research-engine in skills/evolve-solution-population/SKILL.md and read by ahel’s review.

Purpose

Maintain diverse candidate solutions, mutate/recombine them, select under fitness and novelty criteria, and avoid premature convergence.

Input contract

mode_contracts:
  mutation-selection: &evolution_input
    required: [initial_solution_population, fitness_criteria]
    optional: [mutation_operators, novelty_criteria, iteration_budget]
    constraints: [population_members_must_be_comparable_under_declared_criteria]
  novelty-preserving-evolution: *evolution_input

Execution protocol

Do not perform called SOP operations inline; each loaded SOP owns its contract and thresholds.

  1. You MUST load skill mutate-solution-population to mutate or recombine the population.
  2. You MUST load skill select-solution-variants to select solution variants.
  3. You MUST load skill measure-portfolio-diversity to measure portfolio diversity.
  4. You MUST load skill assess-sensitivity to assess selection sensitivity.
  5. You MUST load skill synthesize-idea to synthesize the retained set. If the retained variants require an explicit final ordering, consider rank-candidates as the next tactic. Deviation: stop when added variation no longer changes the frontier or diversity objective; retain niche candidates when explicitly protected.

Mode branches

  • mutation-selection: mutate/recombine, score, and select across iterations.
  • novelty-preserving-evolution: apply diversity pressure and protect useful niches before convergence.

Output contract

mode_contracts:
  mutation-selection:
    produces: [mechanism_to_design_mapping, generated_solutions, synthesis]
    delta_fields: [findings, hypothesis_updates, uncertainties, decisions, open_questions]
  novelty-preserving-evolution:
    produces: [novelty_assessment]
    delta_fields: [findings, hypothesis_updates, uncertainties, decisions, open_questions]

Thresholds and quality gates

  • B: mutation/recombination lineage is traceable; selection criteria are explicit; diversity is measured before convergence.

Failure and counterexamples

Reject a single-winner result when novelty or niche preservation was required, and reject variants with untracked mutations.

Provenance map

  • creative-ideation/systematic-enumeration, evolution-strategy, variation-selection: resolved/concept according to exact v3 lookup.
  • Status: creative-ideation/systematic-enumeration, evolution-strategy resolved; variation-selection concept.

Preserved source criteria ledger

  • Preserve mutation, recombination, selection, diversity pressure, and niche preservation.

Context checkpoint / Delta notes

Append population lineage, fitness changes, diversity deltas, and retained niches.

Signals

GitHub stars
501
Forks
41
Last commit
Sep 2026
Advanced
Catalog kind
skill
Key
evolve-solution-population
Source
github.com/yogsoth-ai/de-anthropocentric-research-engine