agentleFS
LeonardSEOLeonardSEO/ambitious-math-research/skills/lean-verification/SKILL.md

lean-verification

Skill0 starsChanged 33 days agoMITView on GitHub

Assess a supplied Lean proof's actual theorem, build evidence, and transitive axioms. Use for formal verification claims, including statement mismatches and incomplete dependencies; informal proof audit and literature originality are separate tasks.

2 other projects use a version of this.

Adopted by

Most recently taken up by

  • Other versions2 projects use a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 2

    1. epam/ai-dial-chat★ 501same name
    2. matt-w-horn/lean-skills★ 1same name

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • In this repository7 other files it gives its agents.

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:

    Check what the supplied Lean artifact proves, and whether it proves the mathematical statement the user asked about.

    line 8
  2. Distinctive:

    Run the narrow relevant Lean/Lake check and inspect warnings as well as exit status.

    line 14
  3. Distinctive:

    Check meaningful remaining risks: notation or definitions that change the statement, unchecked declarations or kernel-bypass settings, and …

    line 16
  4. Distinctive:

    Use the verification report for a compact record.

    line 28

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 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

Lean Verification

Check what the supplied Lean artifact proves, and whether it proves the mathematical statement the user asked about. Positive conclusions should be stated directly with their exact scope; compilation, agent agreement, and external acceptance are distinct evidence.

Check the target, then its proof

  1. Recover the intended statement independently of the candidate implementation. Compare definitions, quantifiers, hypotheses, implicit arguments, instances, and the elaborated target type. A proof of an implication assuming the desired conclusion is not a proof of the original theorem.
  2. Identify the actual project, compiler version, dependencies, source revision, and target declaration. Use the pinned project environment. Inspect supplied build configuration and imports before executing unfamiliar code. Reuse relevant successful evidence for unchanged artifacts; rebuild when the target or its dependencies changed or the evidence does not cover them.
  3. Run the narrow relevant Lean/Lake check and inspect warnings as well as exit status. If tools are missing, report the source assessment and the exact unexecuted command; do not invent compilation evidence or install dependencies automatically.
  4. Inspect #print axioms TargetDeclaration for the target and its transitive dependencies. sorryAx means incomplete proof. Custom axioms require an explicit justification or a conditional verdict. Accept the standard logical axioms propext, Classical.choice, and Quot.sound without mislabeling them as mathematical gaps. Native-evaluation or compiler-trust dependencies require a clearly stated trust boundary appropriate to the installed version.
  5. Check meaningful remaining risks: notation or definitions that change the statement, unchecked declarations or kernel-bypass settings, and unsupported claims about a whole project based on one checked declaration. Choose further replay or external checking only when the task and risk justify it, using Lean assurance.

A textual search for sorry helps locate unfinished work but does not replace the transitive dependency check. A source tree containing an unrelated unfinished theorem does not make a target that has a complete, checked dependency closure incomplete.

State a usable verdict

  • Formally verified in the stated environment: the actual target matches the intended statement, relevant checks succeeded, and its dependency closure has no unfinished or unsupported mathematical assumptions. Name the check and any disclosed trust extensions.
  • Formally conditional or incomplete: specify the assumed axiom, missing goal, statement mismatch, or unavailable execution evidence. A successful build with sorryAx cannot receive a complete verdict.
  • Candidate rejected: give the concrete contradiction, mismatch, or failed check and a repair obligation.

Separate an analytical proof's established status from its formalization status. A lack of Lean tooling or library support is not an analytical refutation. A genuine logical defect is. Separate correctness, originality, and publication acceptance as well.

Use the verification report for a compact record. When a repair is authorized, repair the responsible statement or proof and repeat only the affected checks. On pause, preserve evidence and the exact next obligation without starting new work.

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.