xintaofei/codeg

Brainstorming turns ideas into fully formed designs and specs through collaborative dialogue, asking clarifying questions and proposing approaches before implementation. Use it before any creative or coding work to align on requirements and get approval.

Citation Management searches Google Scholar and PubMed for papers, extracts accurate metadata, validates citations, and generates properly formatted BibTeX entries. Use it when you need to find papers, verify citation information, convert DOIs to BibTeX, or ensure reference accuracy in scientific writing.

Code Review Reception establishes a verification-first workflow for responding to code review feedback, requiring technical evaluation and clarification before implementation rather than performative agreement. Use it when receiving feedback from humans or external reviewers to avoid blind implementation and ensure changes fit the codebase.

Dispatching Parallel Agents delegates independent tasks to multiple concurrent subagents to investigate and fix issues in parallel. Use it when you have 2+ independent tasks that can be worked on without shared state or sequential dependencies.

Executing Plans loads a written implementation plan, reviews it critically, executes all tasks with verification, and reports when complete. Use it when you have a written implementation plan to execute in a separate session with review checkpoints.

Experimental Design plans experiments and studies before data collection, choosing a design, randomizing, blocking, and laying out treatment combinations so results are interpretable. Use it when assigning subjects to groups, screening factors, or setting up factorial or response-surface designs.

Exploratory Data Analysis performs comprehensive exploratory data analysis on scientific data files in over 200 formats, detecting file type, assessing quality, and generating markdown reports. Use it to understand structure, content, and quality before downstream analysis.

Finishing a Development Branch verifies tests, detects the git environment, and presents merge, pull request, or keep-as-is options to complete development work. It cleans up worktrees and branches after local merges or confirmed discards.

Hypothesis Generation formulates testable hypotheses from observations and data, proposes mechanisms, and designs experiments to test them. Use it for scientific inquiry, literature synthesis, and experimental planning across domains.

Paper Lookup searches 10 academic literature APIs—PubMed, PMC, bioRxiv, medRxiv, arXiv, OpenAlex, Crossref, Semantic Scholar, CORE, and Unpaywall—and returns reproducible results with provenance. Use it to find papers, citations, DOIs, abstracts, full text, open-access PDFs, preprints, or author publications.

Peer Review evaluates scientific manuscripts and grant proposals using structured checklists for methodology, statistics, and reporting standards. Use when writing formal peer reviews with constructive feedback.

Requesting Code Review dispatches a code reviewer subagent with precisely crafted context to catch issues before they cascade. Use it after completing tasks, implementing major features, or before merging to verify work meets requirements.

Scholar Evaluation applies the ScholarEval framework to systematically assess research papers, proposals, and literature reviews across key quality dimensions including problem formulation, methodology, analysis, and writing, with quantitative scoring and actionable feedback.

Scientific Brainstorming guides open-ended research ideation sessions, helping scientists explore interdisciplinary connections, challenge assumptions, and identify research gaps before specific data or observations exist.

Scientific Critical Thinking evaluates scientific claims and evidence quality, applying methods like GRADE and Cochrane Risk of Bias to assess experimental design, statistical validity, biases, and confounding. Use it when reviewing research papers, planning studies, or conducting systematic reviews.

Scientific Schematics generates publication-quality scientific diagrams using Nano Banana 2 AI with Gemini 3.1 Pro Preview quality review. Use it for neural network architectures, flowcharts, biological pathways, and system diagrams without manual drawing.

Scientific Visualization creates publication-ready scientific figures with multi-panel layouts, error bars, significance annotations, and colorblind-safe palettes. Use it when preparing journal submissions with specific formatting requirements for Nature, Science, or Cell.

Skill Writing applies test-driven development to creating and refining AI agent skills. Use it when creating new skills, editing existing ones, or verifying skills work before deployment.

Subagent-Driven Development executes implementation plans by dispatching a fresh implementer subagent per task, reviewing each task, and performing a final whole-branch review. Use it when tasks are independent and you want to stay in the current session.

Systematic Debugging guides a root cause investigation before attempting any fix for bugs, test failures, or unexpected behavior. Use it to avoid symptom fixes and ensure fixes address the underlying issue.

Test-Driven Development enforces writing failing tests before implementation code for any feature, bugfix, or behavior change. It guides a strict Red-Green-Refactor cycle and rejects rationalizations for skipping tests.

Using Git Worktrees sets up an isolated git worktree for feature work, ensuring changes stay separate from the current workspace. Use it when starting feature work or before executing implementation plans, preferring native worktree tools and falling back to manual git worktrees.

Using Superpowers requires invoking relevant skills before any response or action, including clarifying questions, and establishes skill priority, red flags, and platform adaptation rules. Use it at the start of every conversation in a Superpowers-enabled environment.

Verification Before Completion enforces running verification commands and confirming output before claiming work is complete, fixed, or passing. Use it before committing or creating PRs.

Writing Plans turns specs and requirements into detailed, task-by-task implementation plans for coding agents. Use it before touching code to decompose work into bite-sized TDD steps with exact files, tests, and commits.