ablation-brainstorm
Ablation-Brainstorm systematically removes individual components from a target system and observes resulting changes to map hidden dependencies and structural weaknesses. Use this skill when you need to understand system fragility, discover non-obvious relationships between parts, or generate innovative redesign concepts grounded in what actually breaks when components disappear.
git clone --depth 1 https://github.com/yogsoth-ai/de-anthropocentric-research-engine /tmp/ablation-brainstorm && cp -r /tmp/ablation-brainstorm/skills/ablation-brainstorm ~/.claude/skills/ablation-brainstormSKILL.md
# Ablation Brainstorm Systematically remove components from a system one by one, observe what changes, and use the resulting insights to generate novel ideas. ## State Ledger | Resource | Target | Current | % | |----------|--------|---------|---| | web-search | 20 | 0 | 0% | | web-research | 5 | 0 | 0% | | paper-overview | 20 | 0 | 0% | | paper-search | 12 | 0 | 0% | | paper-research | 5 | 0 | 0% | ## HARD-GATE Cannot exit strategy until ≥80% of each budget line is consumed OR yield targets are met with justification for remaining budget. ## Available Tactics | Tactic | Role | |--------|------| | coverage-analysis | Ensure all components are ablated | | evaluation-filtering | Score and filter ablation-derived ideas | ## Available SOPs | SOP | Role | |-----|------| | ablation-execution | Remove components one by one, record responses | | dependency-identification | Extract dependency graph from ablation results | | enumeration-synthesis | Synthesize ablation insights into idea report | ## Execution Guidance 1. **Decompose**: List all components of the target system 2. **Ablate**: Run ablation-execution — remove each component, record effects 3. **Analyze**: Run dependency-identification to find critical dependencies 4. **Ideate**: For each critical dependency, ask "what if this were replaced/eliminated/redesigned?" 5. **Filter**: Apply evaluation-filtering to rank generated ideas 6. **Synthesize**: Produce structured report via enumeration-synthesis
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
Map system architecture to ablatable units for ablation studies
Design ablation studies to isolate component contributions in ML systems
Remove components one by one from a system, record the response/impact of each removal.