assumption-challenge
# Assumption Challenge This Claude Code skill uses a dedicated subagent to construct adversarial arguments against a specific assumption, identifying its weakest points and proposing alternatives. Use it when evaluating critical decisions that depend on unproven premises, particularly in research, strategy, or risk assessment where challenging core beliefs is essential to identify potential failure modes and improve decision quality.
git clone --depth 1 https://github.com/yogsoth-ai/de-anthropocentric-research-engine /tmp/assumption-challenge && cp -r /tmp/assumption-challenge/skills/assumption-challenge ~/.claude/skills/assumption-challengeSKILL.md
# Assumption Challenge Attacks a specific assumption adversarially — constructing the strongest argument for why it might be wrong, proposing an alternative assumption, and assessing the impact on the overall decision if the assumption fails. ## Execution Spawns a subagent that takes a single assumption and builds the strongest possible case against it. ## Why Subagent - Challenge requires adversarial mindset isolated from the assumption-maker - Each assumption deserves focused, dedicated attack - Isolation prevents the challenger from being "too kind" to assumptions ## HARD-GATE Output must include: - Challenge argument with evidence or reasoning - At least 1 alternative assumption (what if the opposite is true?) - Impact assessment if the assumption is wrong - Confidence in the challenge itself (is this a real threat or theoretical?)
Experiment-specific - summarize the DARE executor's research design into a clean research_result report, forced to write back into the spec file produced by formated-specs.
Experiment-specific - replaces writing-specs, emits DARE's 4-layer call plan as a clean research_graph schema. Last step forces load formated-result.
loss-1 judge - read a sample's full dialogue and decide whether the user simulator semantically enacted its Policy Card. check-blind.
loss-2 judge - pairwise quality comparison across the n rungs within one topic; decide monotonicity and endpoint separation. check-blind, D1-D5 only.
Strategy: 面对异常的最佳解释推理
Remove components one by one, observe system changes to reveal hidden dependencies and generate ideas from structural gaps.
Map system architecture to ablatable units for ablation studies
Design ablation studies to isolate component contributions in ML systems