advocate-construction
The Advocate Construction skill deploys a specialized subagent to build the strongest defensible case for a rejected candidate or counter-position by identifying legitimate strengths, reframing weaknesses constructively, and developing arguments that demand serious intellectual engagement. Use this when reconsidering dismissed ideas, stress-testing consensus views, or ensuring thorough evaluation of alternatives before final decisions.
git clone --depth 1 https://github.com/yogsoth-ai/de-anthropocentric-research-engine /tmp/advocate-construction && cp -r /tmp/advocate-construction/skills/advocate-construction ~/.claude/skills/advocate-constructionSKILL.md
# Advocate Construction Builds the most compelling argument for a position — typically a rejected candidate or counter-thesis. The advocate's job is to find every legitimate strength, reframe weaknesses, and construct a case that demands serious engagement. ## Execution Spawns a subagent with the advocate role. The subagent receives the rejected candidate and convergence context, then constructs the strongest possible case for resurrection or adoption. ## Why Subagent - Role isolation prevents contamination from prior judgments - Advocate must genuinely argue FOR the position without hedging - Separation from critic/judge roles ensures intellectual honesty ## HARD-GATE Output must include: - Explicit thesis statement - >= 3 supporting arguments with evidence - Reframing of at least 1 perceived weakness as a strength - Acknowledgment of genuine weaknesses (steel-manning, not straw-manning)
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