Research

Extract structured failure mode list from raw scenarios or artifact analysis. Produces standardized failure mode records.

Systematically catalog failure modes — generate edge cases, observe failures, cluster by mechanism, identify triggers and frequency.

Generate targeted solutions for each identified failure mode, ensuring every failure has at least one proposed mitigation.

Group observed failures by mechanism (not symptom), identify common triggers per cluster, estimate frequency and severity.

Tactic: Trace upstream causes and downstream effects of each failure mode. Builds multi-level cause-mode-effect chains for systemic understanding.

Build cause-mode-effect chains tracing upstream root causes and downstream cascading effects for each failure mode.

Campaign: Forward-looking failure analysis combining pre-mortem rapid screening with systematic FMEA deep-dive. Core question: If this artifact fails, how will it fail? Methods: Klein Pre-Mortem 2007, AIAG-VDA FMEA 2019, IEC 60812.

DOE thinking: identify factors, define levels, and explore combinations to systematically cover the design space.

Strategy: Systematic factor removal — remove factors one at a time and observe whether the conclusion remains stable, identifying which factors are load-bearing.

Identify independent, dependent, and control variables for an experiment

List all key factors, conditions, and assumptions that support or enable the artifact's conclusion.

Bridge two unrelated thinking matrices via Koestler bisociation. Identify independent frames of reference and force collision to produce creative insight.

Generate boundary and extreme test values for a given parameter dimension to stress-test claims.

Structured data extraction from deep-read papers — produces comparison tables (method, dataset, metrics, results, limitations). Used by systematic-survey and deep-survey.

Understand the user's background comprehensively — technical stack, project experience, research experience, publications, research directions. Allows user to express interest beyond their resume. Execute once only, never re-run.

SOP: 为异常现象生成候选解释列表

Execute the plan by dispatching fresh subagents per task, monitoring status, and collecting results

Transform validated hypotheses into rigorous, executable experiment designs

SOP: generate executable experiment configuration files

Synthesize complete execution report from all results, tests, and reproducibility data

Monitor execution progress, detect anomalies, and report status

Execute a Research Spec step by step, respecting context protocol, deviation limits, and backtrack rules. Supports multi-session recovery.

Orchestrate the excursion sequence — departure into unrelated domain, force-fit discoveries back to problem, launch springboard ideas.

Full 8-stage Gordon-Prince excursion process. Deliberate departure from the problem into unrelated domains, then force-fit discoveries back.