{
  "markdown": "# Academic Research Skills\n\n[![Validate Skills](https://github.com/VincenzoImp/academic-research-skills/actions/workflows/validate.yml/badge.svg)](https://github.com/VincenzoImp/academic-research-skills/actions/workflows/validate.yml)\n[![License: MIT](https://img.shields.io/badge/license-MIT-blue.svg)](LICENSE)\n\nEight agent skills for the full academic research pipeline:\nSOTA → survey → contributions → papers.\n\nDesigned for repositories created by\n[`create-academic-research`](https://github.com/VincenzoImp/create-academic-research)\n(paired with the 0.3.0 scaffold), whose directory structure and formats the\nskills rely on: one root `references.bib`, `sota/papers/<citekey>/` folders,\n`survey/coverage.md`, self-contained `contributions/<slug>/` folders, and\nper-venue `papers/<slug>/` folders. The skills are portable `SKILL.md`\nfiles and work with any Agent-Skills-compatible runtime.\n\n## Install\n\nProject-local copy (done automatically by the create-academic-research\nwizard):\n\n```bash\nnpx -y skills add VincenzoImp/academic-research-skills --skill '*' --copy -y\n```\n\n## The Skills\n\n| Skill | Use it to |\n|---|---|\n| `digest-paper` | digest one paper: PDF + synthesis + bib entry + citation graph + index row, MCP-verified end to end |\n| `explore-sota` | run the SOTA exploration loop: search, citation chasing, triage queue, digestion — autonomous or targeted |\n| `write-survey` | write the full LaTeX survey from all syntheses, or diff-update it when the SOTA changes |\n| `develop-contribution` | create or regularize a badge-compliant, self-contained contribution with its LaTeX report |\n| `write-paper` | pick a venue, frame the story, and write the manuscript from the survey and contribution reports |\n| `package-artifacts` | build the self-contained submission bundle meeting the venue's badge requirements |\n| `manage-submission` | freeze submission rounds, map reviewer concerns, draft rebuttals, apply revisions, camera-ready |\n| `adversarial-review` | review the survey, a contribution, or a paper as a severe but fair reviewer |\n\nEach skill is a `SKILL.md` procedure with optional `references/` files carrying\nthe deeper conventions (bibliography rules, citation chasing, experiment/data/\nethics/reproduction conventions, venue fit, writing rules, and so on).\n\n## Hard Rules Baked In\n\n- A citation exists only if a scholarly MCP lookup produced it; SOTA skills\n  stop when the scholarly sources are unavailable.\n- Digestion is atomic: paper folder, synthesis, bib entry, citation graph,\n  and index row land together or not at all.\n- The survey's create mode reads every synthesis before writing a word.\n- Archives of submitted papers are immutable.\n- Reviewers (adversarial-review) report; they never edit the artifact.\n\n## Validate\n\n```bash\npython3 scripts/validate_skills.py\n```\n\n## Release\n\nTag-driven: update `CHANGELOG.md`, tag `vX.Y.Z`, push the tag.\n",
  "bytes": 2879,
  "sha": "c56af47ca5807c15dd4d8df5e43ffb823460e409b974ddfc08a510af3aa0dbcd",
  "repo_slug": "vincenzoimp/academic-research-skills",
  "fonte": "repo",
  "truncated": false,
  "api": "https://api.agentalog.com/api/listings/skl_vincenzoimp_academic_research_skills_wri_ca5a13f2/readme"
}