{
  "markdown": "# agent-research-skills\n\n31 skills for [Claude Code](https://claude.ai/code) covering the full academic research paper lifecycle — from literature search to slide generation — plus GitHub repository analysis for research topics.\n\nExtracted from 17 GitHub repos studying LLM-agent-driven research automation. See [SKILLS_DESIGN.md](/Users/lingzhi/Code/research-engine/SKILLS_DESIGN.md) for the original design specifications.\n\n## Installation\n\n### One-line install\n\n```bash\nnpx skills add lingzhi227/agent-research-skills -g -a claude-code\n```\n\n> Use the `-g` (global) flag. Scripts use `~/.claude/skills/` paths that require global installation.\n\n### Post-install setup\n\n```bash\ngit clone https://github.com/lingzhi227/agent-research-skills.git /tmp/agent-research-skills\n/tmp/agent-research-skills/install.sh\nrm -rf /tmp/agent-research-skills\n```\n\nThis installs slash commands, checks Python dependencies, and verifies script syntax.\n\n### Optional dependencies\n\n| Package | Required by | Install |\n|---------|------------|---------|\n| Python 3 | All scripts | `brew install python3` |\n| PyMuPDF | `self-review`, `deep-research` (PDF parsing) | `pip install PyMuPDF` |\n| numpy + scipy | `data-analysis` (statistical tests) | `pip install numpy scipy` |\n\n### Optional configuration\n\n1. **Semantic Scholar API key** (higher rate limits for literature search):\n   - Get one at https://www.semanticscholar.org/product/api#api-key\n   - Save in `~/keys.md`: `S2_API_Key: your-key-here`\n\n2. **Output directory**: Deep research outputs go to `~/deep-research-output/` by default.\n\n## Available Skills (31)\n\n### Phase 0: Research Discovery & Planning\n\n| Skill | Description | Scripts |\n|-------|-------------|---------|\n| **[github-research](skills/github-research/)** | 6-phase GitHub repo discovery, analysis, and integration planning for research topics | 13 scripts: search, clone, analyze structure/deps/impls, compare, compile report |\n| **[deep-research](skills/deep-research/)** | 6-phase systematic literature survey (frontier → survey → deep dive → code → synthesis → report) | 7 scripts: search APIs, PDF extraction, paper DB, BibTeX, report compilation |\n| **[literature-search](skills/literature-search/)** | Multi-source academic search (Semantic Scholar, arXiv, OpenAlex, CrossRef) with ranking | 4 scripts: `search_crossref.py`, `download_arxiv_source.py`, `search_openalex.py` + shared |\n| **[literature-review](skills/literature-review/)** | Multi-perspective dialogue simulation with expert personas for grounded literature review | Shares search scripts |\n| **[idea-generation](skills/idea-generation/)** | Generate and score research ideas (Interestingness/Feasibility/Novelty) with iterative refinement | 1 script: `novelty_check.py` |\n| **[novelty-assessment](skills/novelty-assessment/)** | Harsh-critic novelty evaluation with up to 10 rounds of literature search | Shares search scripts |\n| **[research-planning](skills/research-planning/)** | 4-stage research plan design with task dependency graphs | Prompt-only |\n\n### Phase 1: Method Design\n\n| Skill | Description | Scripts |\n|-------|-------------|---------|\n| **[atomic-decomposition](skills/atomic-decomposition/)** | Decompose ideas into atomic concepts with bidirectional math ↔ code mapping | Prompt-only |\n| **[algorithm-design](skills/algorithm-design/)** | Algorithm pseudocode (LaTeX) + UML diagrams (Mermaid) with consistency verification | Prompt-only |\n| **[math-reasoning](skills/math-reasoning/)** | Derivations, proofs, formalization, statistical test selection, notation tables | Prompt-only |\n| **[symbolic-equation](skills/symbolic-equation/)** | LLM-guided evolutionary search for scientific equation discovery | Prompt-only |\n\n### Phase 2: Experiment Pipeline\n\n| Skill | Description | Scripts |\n|-------|-------------|---------|\n| **[experiment-design](skills/experiment-design/)** | 4-stage progressive experiment planning (implement → tune → research → ablate) | 1 script: `design_experiments.py` |\n| **[experiment-code](skills/experiment-code/)** | ML training/evaluation pipeline generation with iterative improvement | Prompt-only |\n| **[code-debugging](skills/code-debugging/)** | Structured error analysis with categorization and 4-retry fix loop | Prompt-only |\n| **[data-analysis](skills/data-analysis/)** | Statistical analysis with 4-round code review and appropriate test selection | 2 scripts: `stat_summary.py`, `format_pvalue.py` |\n\n### Phase 3: Paper Writing\n\n| Skill | Description | Scripts |\n|-------|-------------|---------|\n| **[paper-writing-section](skills/paper-writing-section/)** | Section-by-section writing with section-specific guidance and two-pass refinement | Prompt-only |\n| **[related-work-writing](skills/related-work-writing/)** | Related Work section with thematic organization and compare-and-contrast style | Prompt-only |\n| **[survey-generation](skills/survey-generation/)** | Complete survey paper via RAG-based subsection writing with citation validation | Shares search scripts |\n| **[paper-to-code](skills/paper-to-code/)** | Convert paper PDF to runnable code repo (Planning → Analysis → Coding pipeline) | Prompt-only |\n\n### Phase 4: Figures, Tables & Citations\n\n| Skill | Description | Scripts |\n|-------|-------------|---------|\n| **[figure-generation](skills/figure-generation/)** | Publication-quality matplotlib figures with VLM feedback loop (10 figure types) | 1 script: `figure_template.py` |\n| **[table-generation](skills/table-generation/)** | JSON/CSV → LaTeX booktabs tables with bold-best, significance stars, multi-dataset | 1 script: `results_to_table.py` |\n| **[citation-management](skills/citation-management/)** | BibTeX harvesting, validation, deduplication, and auto-fix | 2 scripts: `harvest_citations.py`, `validate_citations.py` |\n| **[backward-traceability](skills/backward-traceability/)** | Every PDF number hyperlinks to the code line that produced it | 1 script: `ref_numeric_values.py` |\n\n### Phase 5: LaTeX & Compilation\n\n| Skill | Description | Scripts |\n|-------|-------------|---------|\n| **[latex-formatting](skills/latex-formatting/)** | Conference templates (ICML/ICLR/NeurIPS/AAAI/ACL), formatting fixes, pre-submission checks | 2 scripts: `latex_checker.py`, `clean_latex.py` |\n| **[paper-compilation](skills/paper-compilation/)** | Full pdflatex+bibtex pipeline with auto-fix error correction loop | 2 scripts: `compile_paper.py`, `fix_latex_errors.py` |\n| **[excalidraw-skill](skills/excalidraw-skill/)** | Programmatic Excalidraw diagramming via MCP tools with quality verification | MCP server (7 CJS files) |\n\n### Phase 6: Review & Polish\n\n| Skill | Description | Scripts |\n|-------|-------------|---------|\n| **[self-review](skills/self-review/)** | 3-persona automated review (NeurIPS form) with reflection and meta-review | 2 scripts: `extract_pdf_text.py`, `parse_pdf_sections.py` |\n| **[paper-revision](skills/paper-revision/)** | Map reviewer concerns to sections, apply targeted edits, verify improvements | Prompt-only |\n| **[rebuttal-writing](skills/rebuttal-writing/)** | Point-by-point rebuttal with evidence-based responses | Prompt-only |\n| **[slide-generation](skills/slide-generation/)** | Paper → Beamer slides (extract elements, generate skeleton, simplify) | 1 script: `extract_paper_elements.py` |\n| **[paper-assembly](skills/paper-assembly/)** | End-to-end pipeline orchestrator with 9-phase checkpointing | 1 script: `assembly_checker.py` |\n\n## Usage\n\n### Slash commands\n\n```\n/research transformer architectures for long-context reasoning\n```\n\n### Natural language (skills activate automatically)\n\n```\n\"Analyze GitHub repos for multi-agent coordination research\"\n\"Do a literature review on protein folding with LLMs\"\n\"Write the Methods section of my paper\"\n\"Generate a comparison table from results.json\"\n\"Review my paper draft before submission\"\n\"Make slides from my paper\"\n\"Check if my idea is novel\"\n\"Design experiments for my contrastive learning method\"\n```\n\n### Direct script usage\n\n```bash\n# Search GitHub repos for a research topic\npython ~/.claude/skills/github-research/scripts/search_github.py --query \"multi-agent LLM coordination\" --max-results 50 --output repos.jsonl\n\n# Analyze a cloned repo's structure\npython ~/.claude/skills/github-research/scripts/analyze_repo_structure.py --repo-dir ./my-repo --output analysis.json\n\n# Search literature\npython ~/.claude/skills/literature-search/scripts/search_crossref.py --query \"attention mechanism\" --rows 10\n\n# Validate citations\npython ~/.claude/skills/citation-management/scripts/validate_citations.py --tex paper/main.tex --bib paper/references.bib\n\n# Generate experiment design\npython ~/.claude/skills/experiment-design/scripts/design_experiments.py --method \"contrastive learning\" --task classification --format markdown\n\n# Format p-values\npython ~/.claude/skills/data-analysis/scripts/format_pvalue.py --values \"0.001 0.05 0.23\" --format latex\n\n# Extract paper elements for slides\npython ~/.claude/skills/slide-generation/scripts/extract_paper_elements.py --tex main.tex --output slides.tex\n\n# Check paper pipeline completeness\npython ~/.claude/skills/paper-assembly/scripts/assembly_checker.py --dir paper/ --verbose\n```\n\n## Architecture\n\nAll skills follow the same structure:\n\n```\nskills/<skill-name>/\n├── SKILL.md           # Skill definition (prompt, workflow, rules)\n├── scripts/           # Executable tools (optional)\n│   └── *.py           # CLI scripts with argparse, docstring headers\n└── references/        # Reference docs (optional)\n    └── *.md           # Templates, API docs, patterns\n```\n\n**Design principles:**\n- Scripts are stdlib-only where possible (no heavy dependencies)\n- Every script has `--help`, docstring header with usage examples, and argparse CLI\n- Skills link to each other via `## Related Skills` sections (upstream/downstream/see-also)\n- Prompt-only skills are fully self-contained in SKILL.md\n\n## Requirements\n\n- [Claude Code](https://claude.ai/code)\n- Python 3\n- Optional: PyMuPDF, numpy, scipy (see installation)\n",
  "bytes": 10016,
  "sha": "a67283e20e50045e1949c134900431113b357b5d216042a049b679bda5bdae69",
  "repo_slug": "lingzhi227/agent-research-skills",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/skl_lingzhi227_agent_research_skills_table_g_b123fe99/readme"
}