limits-colimits — Claude Code skill: what it does, how to install
limits-colimits is a Claude Code skill that provides decision workflows for category theory limits and colimits, including Lean 4 proof formalization and SymPy calculations.
MIT ★ 3,951 updated 2026-01-26
npx skills add parcadei/Continuous-Claude-v3 --skill limits-colimits -a claude-codeWhat limits-colimits does
This skill guides Claude Code through category theory problems. It provides a decision tree to identify limit types like equalizers, pullbacks, and products, along with their dual colimit counterparts such as pushouts and coequalizers.
It outlines specific workflows for Lean 4 using the CategoryTheory.Limits namespace and universal property helpers like IsLimit.lift. For concrete sets, the skill triggers symbolic equation solving via scripts/sympy_compute.py using uv run.
Working with this skill requires lake for Lean builds and an execution environment with uv and Python for the SymPy script.
What it helps with
- classify diagram types into limits or colimits
- formalize universal cone properties using Lean 4
- compute pullback equations with SymPy
- verify limit preservation under right adjoint functors
Example prompts for the agent
- Help me formulate the universal property of this pullback in Lean 4
- Solve the equalizer condition f(a) == g(b) using the SymPy harness
- Run lake build to verify my limit preservation proof
How to install limits-colimits
1Skills CLI (Vercel)
npx skills add parcadei/Continuous-Claude-v3 --skill limits-colimits -a claude-code2Manually: put the files in place
~/.claude/skills/limits-colimits/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/limits-colimits/
1Skills CLI (Vercel)
npx skills add parcadei/Continuous-Claude-v3 --skill limits-colimits -a codex2Manually: put the files in place
.agents/skills/limits-colimits/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add parcadei/Continuous-Claude-v3 --skill limits-colimits -a cursor2Manually: put the files in place
.cursor/skills/limits-colimits/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install parcadei/Continuous-Claude-v3 .claude/skills/math/category-theory/limits-colimits2Manually: put the files in place
.github/skills/limits-colimits/SKILL.md1Skills CLI (Vercel)
npx skills add parcadei/Continuous-Claude-v3 --skill limits-colimits -a gemini-cli2Platform CLI
gemini skills install https://github.com/parcadei/Continuous-Claude-v3 --path .claude/skills/math/category-theory/limits-colimitsManually: put the files in place
.devin/skills/limits-colimits/SKILL.mdThe legacy .windsurf/skills/ path still works
Claude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about limits-colimits
How to install limits-colimits in Claude Code?
Run npx skills add parcadei/Continuous-Claude-v3 --skill limits-colimits -a claude-code. The skill file will be placed in ~/.claude/skills/limits-colimits/SKILL.md.
What does limits-colimits do?
It provides a step-by-step reasoning tree for categorical limits and colimits, bridging abstract diagrams to Lean 4 constructs and SymPy solvers.
Which tool permissions are required?
The skill requests Bash and Read permissions to inspect files, execute SymPy calculations via uv, and run lake build.
Source text
SKILL.md · MIT
---
name: limits-colimits
description: "Problem-solving strategies for limits colimits in category theory"
allowed-tools: [Bash, Read]
---
# Limits Colimits
## When to Use
Use this skill when working on limits-colimits problems in category theory.
## Decision Tree
1. **Identify Limit Type**
- Product: limit of discrete diagram
- Equalizer: limit of parallel pair f, g: A -> B
- Pullback: limit of A -> C <- B
- Terminal object: limit of empty diagram
- Lean 4: `CategoryTheory.Limits` namespace
2. **Verify Universal Property**
- Cone from L with projections pi_i: L -> D_i
- For any cone from X, unique morphism u: X -> L
- Triangles commute: pi_i . u = cone_i
- Lean 4: `IsLimit.lift` gives the unique morphism
3. **Colimit (Dual)**
- Coproduct: colimit of discrete diagram
- Coequalizer: colimit of parallel pair
- Pushout: colimit of A <- C -> B
- Initial object: colimit of empty diagram
4. **Compute Limits Concretely**
- In Set: product = Cartesian product
- Equalizer = {x | f(x) = g(x)}
- Pullback = {(a,b) | f(a) = g(b)}
- `sympy_compute.py solve "f(a) == g(b)"`
5. **Preservation**
- Right adjoint preserves limits
- Left adjoint preserves colimits
- Representable functors preserve limits
- Lean 4: `Adjunction.rightAdjointPreservesLimits`
- See: `.claude/skills/lean4-limits/SKILL.md` for exact syntax
## Tool Commands
### Lean4_Limit
```bash
# Lean 4: import CategoryTheory.Limits.Shapes.Products
```
### Lean4_Universal
```bash
# Lean 4: IsLimit.lift cone -- unique morphism from universal property
```
### Sympy_Pullback
```bash
uv run python -m runtime.harness scripts/sympy_compute.py solve "f(a) == g(b)"
```
### Lean4_Build
```bash
lake build # Compiler-in-the-loop verification
```
## Cognitive Tools Reference
See `.claude/skills/math-mode/SKILL.md` for full tool documentation.
© parcadei · MIT · parcadei/Continuous-Claude-v3