Steven Denney

Draft a referee report on someone else's manuscript for a journal editor, using five adversarial finders, a Blue Team error filter, a Chief Reviewer synthesis, and a Tone Guard legal pass. Use when the user has been asked to referee a manuscript; self-audits go to paper-review-lite or presubmit, author responses to referee-response.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

Turns a raw Qualtrics conjoint export into analysis-ready long-format data: implementation identification, reshaping, choice-profile alignment, reference categories, and pilot exclusion. Use when a conjoint CSV needs cleaning, reshaping, or debugging before AMCE estimation.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

Designs multi-country survey experiments: case selection, localization and TRAPD translation, stimulus calibration, per-country power, and pooled versus per-country analysis with measurement-equivalence checks. Use when fielding the same experiment in several countries.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

Reviews a whole deliverable against its manifest without rewriting it — deterministic gate, per-section finders, a whole-work pass, and adjudicated findings.json and report.md with honest coverage. Use before shipping a deliverable or for the night shift.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

Read or convert documents (PDF, Office, OpenDocument, RTF, EPUB, CSV) in a research workflow: read directly or convert, pick the converter from document structure, and decide whether the Markdown is tracked or scratch. Use when a document must be read, quoted, summarized, searched, extracted, or added to a source library; scans go to $vlm-ocr.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

Turns a theory into falsifiable, pre-registerable hypotheses: DGP and DAG, identification, estimands, SESOIs, test choice among NHST, equivalence, and minimum-effect tests, scope conditions, and hypothesis tiers. Use when converting a research question into testable predictions or predicting a null.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

Designs and diagnoses list experiments (the item count technique): whether indirect measurement is warranted, control-list construction, design variants, estimators and assumption tests in the list package, and power. Use when measuring a sensitive item or debugging an implausible prevalence estimate.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

Builds or audits a literature review: evidence map, closest prior work, gap verdict, and synthesis plan. Use to check whether a contribution is novel, map what a literature establishes, or organize a review; systematic reviews get a protocol scaffold.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

Reads a model's own uncertainty from token log-probabilities: collecting and aggregating label logprobs, confidence tiers and margins for triage, calibration against human labels (ECE, Brier, reliability diagrams), downstream use, and reproducibility. Use for per-item classifier confidence or routing low-confidence cases to human review; codebook and validation design go to text-classification.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

Orchestrate complex work from an active gpt-6-astra or gpt-6.1-sol Codex session. The detected lead owns decomposition, integration, and verification; Astra keeps the hardest reasoning in the lead, while Sol escalates unusually difficult units to Astra. Both route bounded work to Sol at medium effort, mechanical work to Luna, and can use a cross-vendor Claude peer. Not for routine single-context work. Use when the user explicitly asks to orchestrate, delegate, fan out, parallelize, assign subagents, obtain independent checks, or have Codex act as tech lead.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

Organize an author's response to peer review: extract every referee point, order the revision by dependency, flag possible pushbacks, and build the response-letter skeleton the author fills in. Use for revise-and-resubmit planning, response letters, or checking every point was answered.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

Scaffold or audit a social-science replication package, and audit the manuscript and archived research objects against FAIR. Use when setting up or repairing a package, checking one before submission, or drafting data, code, and materials availability statements. Adapted from Yusaku Horiuchi's replication-package-guide.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

Scaffold or audit a research-project repository organized around its source library (PDFs, Markdown conversions, bibliography, intake pipeline) plus the analysis, manuscript, and review folders it needs. Use when starting or auditing a research repo. Not for one-PDF intake or publication replication packages.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

Plans a research project as a durable decision map, resolving one design decision per session until the design is defensible and pre-registerable. Use at project start, when a design has more open decisions than one conversation holds, or when planning keeps restarting. Not for running analyses or writing the paper.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

Audit fielded survey response data for registered elements, data quality, bot and AI-automation screening, and sample integrity, and emit an appendix-ready quality report. Use at soft launch or after fielding, before outcomes are analyzed.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

Advises on setting up and validating topic models for survey and experimental text: discovery vs. measurement, STM/LDA/BERTopic choice, preprocessing, covariate formulas, choosing K across coherence, FREX, held-out likelihood, and residuals, validating topics against human coding, robustness, and reporting. Coding against a fixed codebook goes to text-classification.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

OCR scanned or image-only corpora with vision-language models in four phases: `triage` inventories the scans, identifies script and language, and shortlists models by script, hardware, and data constraints; `evaluate` picks a system by measured CER/WER against ground truth, using born-digital twins, a second reader, and decision rules set in advance; `run` builds the production pipeline; and `clean` removes model artifacts, audits what every cleaning rule deletes, corrects with rule-based and LLM passes, and records provenance. Born-digital documents with a text layer go to $doc-to-markdown; runs on Leiden's ALICE cluster add $vlm-ocr-alice.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

Consults Fable as an independent, read-only reviewer at an effort the caller matches to the question. The calling session (usually Opus, or Haiku) holds the task and escalates one decision point. Use before committing to an interpretation or a substantial piece of writing or analysis, when stuck or not converging, when changing approach, or as a final check on a finished task. Fallback for when the native advisor tool is unavailable. Not for a session already running Fable.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

Designs conjoint and factorial-vignette experiments, covering attribute architecture and randomization restrictions, effective-N power, treatment realism, the choice among AMCE, marginal-mean, and AMIE estimands, forced-choice versus rating designs, and PAP tiers. Use when the user is planning a conjoint, drafting or critiquing an attribute table, asking how many respondents or tasks are needed, whether to report AMCEs or marginal means, or how to test interactions. Reviewing an existing design goes to conjoint-diagnostics, cleaning the export to conjoint-cleaning.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

Reviews an existing conjoint study for threats to inference and returns severity-ranked findings on design integrity, estimation, measurement error, external validity, and interpretation, including the guardrail against reading an AMCE as majority preference. Use when the user asks whether a conjoint design or analysis holds up, has referee comments on a conjoint, or wants a second opinion on estimation or interpretation. New designs go to conjoint-design, reshaping to conjoint-cleaning.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

Designs multi-country survey experiments, covering case selection with a stated comparative logic, instrument localization and TRAPD translation, stimulus calibration, per-country power, and pooled versus per-country analysis with measurement-equivalence checks. Use when the user fields the same experiment in several countries, asks which countries to include, how to adapt an instrument without breaking comparability, whether to pool, or how large each country sample must be. Question wording goes to survey-design, reporting to methods-reporting, the study plan to pre-registration-writing.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

Generates 3-5 conceptually distinct approaches labeled Novel, Surprising, Diverse, or Conventional, and holds for the user's selection before implementing. The --codex mode runs the brainstorm on GPT-6 Astra via codex exec, then has Codex implement the selected approach. Use before committing to an approach when a creative, architectural, or analytical task has more than one non-obvious solution; use --codex when another model family should widen the range.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

Designs conjoint and factorial-vignette experiments: attributes and restrictions, effective-N power, estimand choice among AMCE, marginal means, and AMIE, forced choice versus rating, and PAP tiers. Use when planning a conjoint, critiquing an attribute table, sizing respondents and tasks, or testing interactions.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

Turns dictated captures in a deliverable's inbox/ into a proposed planning-wiki diff, classifying and tracing each statement; decisions need the author's assent. Use to intake notes or process the inbox.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.