Loading...
Loading...
Found 57 Skills
Decompose research ideas into atomic, self-contained concepts with bidirectional math-code mapping. For each concept, extract the math formula from papers and find code implementations. Use for complex system papers requiring formal grounding.
Research a topic in depth using web search, academic papers, and citation graphs. Use when the user asks to research, investigate, or explore a topic thoroughly.
Research topics with web search. Use when: researching a topic or concept, finding current information, answering factual questions, comparing options or technologies. Triggers: research [topic], find out about, what are the best practices for, research the latest on.
Workflow 1: Full idea discovery pipeline. Orchestrates research-lit → idea-creator → novelty-check → research-review to go from a broad research direction to validated, pilot-tested ideas. Use when user says "找idea全流程", "idea discovery pipeline", "从零开始找方向", or wants the complete idea exploration workflow.
Orchestrates end-to-end autonomous AI research projects using a two-loop architecture. The inner loop runs rapid experiment iterations with clear optimization targets. The outer loop synthesizes results, identifies patterns, and steers research direction. Routes to domain-specific skills for execution, supports continuous agent operation via Claude Code /loop and OpenClaw heartbeat, and produces research presentations and papers. Use when starting a research project, running autonomous experiments, or managing a multi-hypothesis research effort.
Search, filter, and format entries from BibTeX or BibLaTeX .bib files for research workflows. Use when a user wants to find papers, search a bibliography, filter a library, or look up references by topic, author, year, venue, DOI, arXiv ID, keywords, annotation, abstract, or entry type. Handles Zotero-exported libraries. Supports compact search expressions such as author:, year-gte, type:, and has:, combined filters, research-oriented output fields, raw BibTeX export, and LaTeX/Typst citation snippet generation.
Sync verified experiment results from the code repo or a code worktree into the paper's daily experiments log and project memory. Use when results in code/docs/results, code/docs/reports, code/docs/runs, worktree docs, logs, or user-confirmed metrics should be promoted into paper-facing evidence.
Create a new Git branch or code worktree for experiments, features, baselines, rebuttal fixes, or method revisions. Use when starting an isolated code direction, creating a branch, creating a project-aware code worktree under a project control root, or setting up a worktree with UV sync, IDE config copying, linked assets, and worktree memory.
Diagnose surprising, negative, unstable, or ambiguous ML/AI experiment results and decide whether to debug implementation, rerun experiments, change metrics or baselines, revise the algorithm, narrow the paper claim, park, or kill a direction. Use this skill whenever results do not match expectations, a method fails, metrics conflict, seeds vary, baselines beat the method, plots look suspicious, or the user asks what to do next after experimental results.
Conducts citation-backed research using Firecrawl MCP search, scrape, map, crawl, and agent tools with selectable quick, standard, deep, and ultradeep modes. Use for multi-source comparisons, technical evaluations, market research, and high-stakes decision support.
Structured web research workflow that ensures research results are incrementally saved to files, preventing loss due to session truncation. This skill is triggered when users say "research", "search for information", "help me look up", "learn about", or "latest information".
External research workflow for docs, web, APIs - NOT codebase exploration