destructive-ideation

SkillDev tools

Lets your agent break out of stuck thinking by challenging assumptions and generating new problem-solving angles.

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 destructive-ideation skill

About this skill

Break fixation by surfacing assumptions and applying deliberate provocation modes (reversal, negation, random entry, exaggeration, extreme/zero-resource constraint, distortion), then extract constructive movement and synthesize candidate directions.

What this skill tells your AI

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

Purpose

Break fixation with explicit provocation modes, then extract constructive movement and candidate directions.

Input contract

mode_contracts:
  reverse: &destructive_input
    required: [current_frame, target_problem]
    optional: [load_bearing_assumptions, constraints]
    constraints: [provocation_must_name_the_frame_or_assumption_it_attacks]
  negation: *destructive_input
  random-entry: *destructive_input
  extreme-constraint: *destructive_input
  sacred-cow: *destructive_input
  distortion: *destructive_input

Execution protocol

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

  1. You MUST load skill surface-assumptions to surface assumptions.
  2. You MUST load skill generate-provocation to generate the mode-specific provocation.
  3. You MUST load skill extract-constructive-movement to extract constructive movement.
  4. You MUST load skill synthesize-idea to synthesize an idea.
  5. You MUST load skill build-concept-fan to expand the concept fan. If the idea now requires explicit component transformation, consider structural-transformation. If it exposes a technical or physical contradiction, consider resolve-inventive-contradiction. Deviation: mode may be reverse, negation, random-entry, extreme-constraint, sacred-cow, or distortion; record the selected mode.

Mode branches

  • reverse: invert the current desired/undesired relation.
  • negation: negate a load-bearing assumption or axiom.
  • random-entry: introduce an explicitly recorded external/random concept.
  • extreme-constraint: force zero-resource or boundary conditions.
  • sacred-cow: challenge an explicitly protected convention.
  • distortion: exaggerate, compress, or otherwise perturb a salient feature.

Output contract

mode_contracts:
  reverse: &assumption_attack_output
    produces: [assumption_targets, provocations, constructive_movements, idea_set]
    delta_fields: [findings, hypothesis_updates, assumption_updates, decisions, recommended_jumps]
  negation: *assumption_attack_output
  random-entry: &provocation_output
    produces: [provocations, constructive_movements, idea_set]
    delta_fields: [findings, hypothesis_updates, assumption_updates, decisions, recommended_jumps]
  extreme-constraint: *assumption_attack_output
  sacred-cow: *assumption_attack_output
  distortion: *provocation_output

Thresholds and quality gates

  • B: selected mode is explicit; each provocation has a constructive movement; candidate ideas retain a trace to the attacked frame.

Failure and counterexamples

Reject unbounded randomness, provocation without a target, and ideas that cannot state what constructive movement they preserve.

Provenance map

  • assumption-destruction, axiom-negation, provocation-and-movement, provocation-generation, movement-extraction: resolved/concept per exact v3 lookup.
  • Status: assumption-destruction, axiom-negation, provocation-and-movement, movement-extraction resolved; provocation-generation concept (only package-prefixed variants found).

Preserved source criteria ledger

  • Preserve reversal, negation, random entry, exaggeration/extreme constraint, distortion, and movement extraction.

Context checkpoint / Delta notes

Append mode, attacked assumptions, provocations, movements, and selected directions.

Signals

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