agentleFS
VeryMathVeryMath/AI4Math-Paper-Reading/skills/math-paper-reading/SKILL.md

math-paper-reading

Skill6 starsChanged 4 months agoMITView on GitHub

Read mathematical papers from their source text, using correctable structure indexes for theorem extraction, dependency analysis, proof explanations, literature queries, and reference notes.

The file

We split this skill into the rules it gives an agent

4 rules. About 0 are common, 4 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Start with agent_router.md and load only the modules needed for the user's current task.

    line 8
  2. Distinctive:

    Use the original paper to check what its authors state, assume, and cite.

    line 31
  3. Distinctive:

    Keep source quotations separate from interpretation and Agent-proposed derivations.

    line 38
  4. Distinctive:

    Use available code tools or existing functions for exact graph calculations.

    line 44

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Math Paper Reading

Start with agent_router.md and load only the modules needed for the user's current task. A full-paper extraction is optional for a local question or a targeted literature query.

Resolve module references relative to this package. Read papers from their actual supplied locations and save working outputs in the user's project; do not treat the package's examples or reference database as user data.

Modules

  • skill_base.md: extract statements, assumptions, locations, and source-supported dependencies into a working JSON index.
  • skill_pathway_proof.md: compute dependencies and reading paths; keep confirmed, uncertain, and external edges distinct.
  • skill_paper_deep_read.md: explain the paper and separate source text, additional derivations, and unresolved proof steps.
  • skill_reference_manager.md: maintain reference metadata and user notes without treating extracted claims as verified facts.
  • skill_literature_search.md: build source-linked queries and distinguish proposed searches from retrieved evidence.

Source and correction rules

  • Use the original paper to check what its authors state, assume, and cite. JSON is a correctable index, not mathematical ground truth.
  • Consult the relevant source passage whenever a statement, condition, or edge is uncertain or disputed, even when a JSON index already exists.
  • Correct source-supported extraction mistakes in the current index. Explain the source location and the correction, then update affected reports and graph calculations. Do not require entity locks or a full workflow restart.
  • Keep source quotations separate from interpretation and Agent-proposed derivations. Retain unresolved gaps and suspected author errors explicitly; do not silently repair a quotation or declare proof completion.
  • For a confirmed dependency, B.dependencies containing A means the graph edge A -> B. Rebuild cited_by from these confirmed edges. Uncertain or external references do not contribute to confirmed internal graph degrees.
  • Use available code tools or existing functions for exact graph calculations; do not require generating a helper script merely to proceed.
  • Preserve user notes and state the checked scope. Package examples are not user data; graph consistency is not a proof of mathematical correctness.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.