Light Skills
A skill pack for research and competition projects: from idea and literature to experiments, paper and LaTeX typesetting
- 939 ★ on the repository
- Last commit
- MIT
Extended access
Why this level
- Runs local Python and R scripts and assembles project files
- Some skills reach the network for literature through the agent
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/light-skillsDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- Runs local Python and R scripts and assembles project files
- Some skills reach the network for literature through the agent
This is third-party code. Review the repository files before installing.
What it does
Light Skills is a skill pack that carries a research or innovation project from a vague idea to a checkable result. An orchestrator asks questions, splits the work into stages and sets points where a human decision is needed. Separate skills handle literature search, idea critique, experiment planning, data engineering, result analysis, paper writing, citation checks and LaTeX typesetting. Figures are generated programmatically in Python or R so they are reproducible. When a fact or a tool is missing, a skill marks it as unverified rather than inventing it.
Who it is for. For researchers, graduate and undergraduate students and competition entrants who take a project to a paper or a defense.
Good fit when
- You only have a direction and need it turned into a plan and stages
- You need to check an idea's novelty and find strong counterarguments
- You need to write an English paper and typeset it in LaTeX
Not a fit when
- You need a quick draft without fact checks or reproducibility
- You have no Python, R or LaTeX where a given skill needs them
Example request
Help turn this direction into a research plan: ask the needed questions, split it into stages and flag the risksLimitations
Some skills need local tools: Python, R for figures, LaTeX for typesetting. When a tool is missing, a skill honestly marks the step as not done and does not pretend it is complete. Part of the documentation is in Chinese. The public version needs no MCP or local database, facts are checked at task time.
How to disable. Remove the skill folders from your agent's skills directory, for example ~/.claude/skills.
FAQ
Do I need a paid service or a key?
The public version needs no MCP or local database. Some steps need local tools such as Python, R and LaTeX.
What happens if a tool is missing?
The skill marks the step as not done or unavailable and offers a choice, but never passes a result off as complete.
Discussion
No questions or reviews yet.
Related
SkillsA skill set for research work in the style of Nature journals: finding and reading papers, writing, statistics, figures, reviewer responses
Skill with 30 prompt templates for research papers in English and Chinese, from literature review to grant proposals
A skill set for academic papers: deep method analysis, visual slides and a coherent write-up from a paper PDF
Research Skills gives Claude Code, Codex and Copilot CLI plugins for research: literature, grants, papers, figures, slides.
Ten research skills for Claude Code and Codex: figures, conference and journal papers, grants, peer review and code release
Skills tailored to specific academic journals that help choose a venue, prepare a paper and pass review
README in short
The README presents Light Skills as an open, domain-agnostic skill pack for research, competition and innovation projects. It stresses discipline: read before writing, mark unverified items as unknown, build figures programmatically, and ask the human at key forks. Install is through a bootstrap script or by copying into the skills directory of Claude Code, Codex or OpenCode. It includes end-to-end project examples and environment requirements for LaTeX and R.
README badge
Original SKILL.md
--- name: light-orchestrator description: Coordinate and recover multi-stage Light research projects through the canonical .light/passport.yaml state, stages 1-13, resident overlays, checkpoints, findings, parallel joins, stale propagation, handoffs and user-authorized reroutes. Use for a new/resumed/partial/dirty/failed/stale/delivered research project; when the user says continue, resume, take over, checkpoint, reroute, recover or deliver; or when work crosses two or more Light research stages. --- # Light orchestrator Coordinate, route and recover the research lifecycle. Do not impersonate the stage skills and do not turn a deterministic check into a research judgment. ## Non-negotiable boundaries 1. Never choose the research direction, final idea, plan, venue, back-edge, revision-budget exception, known-limitation conversion or final delivery for the user. Present a recommendation, evidence, alternatives and consequences, then stop. 2. reroute.py is advisory. Only passport.py add-back-edge may write a real back-edge, and only after the user authorizes that exact route.