arc

ARC

Agent Research Copilot (ARC) is an angentic research toolkit for theoretical physics knowledge domain construction, idea generation and calculation workflows. It works as a plugin of coding agents such as Codex / Claude Code, with the strength of bringing coding agents into a research context, and generating publication-level ideas in theoretical research.

ARC is CLI-first. Its workflow Skill uses six core Python commands; a seventh, separately installed adapter exposes the same services through optional MCP:

• arc-paper: paper metadata, references, citers, ar5iv sections, equation context, full-text search, LLM paper summaries, and paper-summary batches.
• arc-domain: builds a cached research-domain package from a seed paper and optional scientific intent.
• arc-llm: reusable host LLM execution, provider selection, and proposers-reviewer workflows.
• arc-typeset: deterministic typesetting utilities, including Markdown to PDF conversion through Pandoc and XeLaTeX.
• arc-companion: builds a source-faithful original/translation/commentary PDF from INSPIRE/ar5iv papers with a unified glossary and deterministic LaTeX.
• arc-jobs: protocol-neutral persistent background execution for ARC CLIs.
• arc-mcp (optional): exposes ARC services to MCP clients; it delegates background work to arc-jobs.
• plugins/arc/skills/arc: agent-facing workflow instructions for domain building, idea generation, and research calculations.

Who This Is For

Use ARC when you want to:

• Look up reliable paper metadata, references, citers, sections, or equations.
• Summarize a paper from cached ar5iv/INSPIRE data.
• Generate a Chinese-by-default companion-reading PDF with a glossary and an original, translation, companion sequence while retaining the paper's equations, equation numbers, figures, tables, and bibliography.
• Build a research-domain overview from a seed paper.
• Generate ideas using domain context and reviewer scoring.
• Plan and execute a careful symbolic or numerical research calculation with explicit provenance and checks.

Deterministic paper queries do not need an LLM. Paper summaries, domain briefings, idea loops, and calculation workflow runners need a host LLM provider.

Install

Remarks:

• Permission: the same as many heavy skills/plugins, ARC will need permissions to run Python scripts. Accepting permissions could be annoying. We recommend installing ARC within docker or a virtual machine, and allow all permissions in that virtual environment. As always for working with AI agents, be aware of risk to your data and system.

• Token usage. As measured using Claude + DeepSeek, a typical run of domain build + idea generation consumes about 1M uncached input tokens, and 0.5M output tokens, in about an hour's running time. The token usage may vary depending on the specific tasks and LLM used. Be aware of token usage and costs.

• If ARC has played a role in your research, please consider citing the ARC manual.

Citation

Yanjiao Ma, Yi Wang, and Xingkai Zhang. ARC: An LLM-Native Agent Workflow for Theoretical Physics Research. ChinaXiv:202606.00234, 2026. https://chinaxiv.org/abs/202606.00234

@misc{ma2026arc,
  title         = {{ARC}: An {LLM}-Native Agent Workflow for Theoretical Physics Research},
  author        = {Ma, Yanjiao and Wang, Yi and Zhang, Xingkai},
  year          = {2026},
  month         = jun,
  publisher     = {ChinaXiv},
  eprint        = {202606.00234},
  archivePrefix = {ChinaXiv},
  url           = {https://chinaxiv.org/abs/202606.00234},
  note          = {Version 1}
}

Requirements:

• Python 3.11 or newer.
• uv for fast first-time CLI runtime setup; Python venv + pip is the fallback.
• Network access for first-time INSPIRE/ar5iv fetches.
• Codex or Claude Code for host LLM work; unknown hosts fall back to manual prompt handoff.
• Optional for arc-typeset md2pdf: pandoc, xelatex, and a CJK-capable font such as Noto Sans CJK SC.
• For arc-companion build: latexmk, xelatex, Poppler command-line tools, and fonts covering the source and annotation languages.

Agent Plugin Setup

ARC can be installed as a host plugin from this repository. plugins/arc/ is the plugin root for both Codex and Claude Code, and plugins/arc/skills/arc/ is the single canonical Skill source. The base plugin contains no MCP manifest or MCP dependency.

Install for Codex (run in shell, or in Codex with ! prefix):

codex plugin marketplace add tririver/arc --ref stable
codex plugin add arc@arc

Install for Claude Code (run in Claude Code):

/plugin marketplace add tririver/arc@stable
/plugin install arc