sheaf-cohomology
Čech cohomology for local-to-global consistency verification in code structure and data schemas.
Best use case
sheaf-cohomology is best used when you need a repeatable AI agent workflow instead of a one-off prompt.
Čech cohomology for local-to-global consistency verification in code structure and data schemas.
Teams using sheaf-cohomology should expect a more consistent output, faster repeated execution, less prompt rewriting.
When to use this skill
- You want a reusable workflow that can be run more than once with consistent structure.
When not to use this skill
- You only need a quick one-off answer and do not need a reusable workflow.
- You cannot install or maintain the underlying files, dependencies, or repository context.
Installation
Claude Code / Cursor / Codex
Manual Installation
- Download SKILL.md from GitHub
- Place it in
.claude/skills/sheaf-cohomology/SKILL.mdinside your project - Restart your AI agent — it will auto-discover the skill
How sheaf-cohomology Compares
| Feature / Agent | sheaf-cohomology | Standard Approach |
|---|---|---|
| Platform Support | Not specified | Limited / Varies |
| Context Awareness | High | Baseline |
| Installation Complexity | Unknown | N/A |
Frequently Asked Questions
What does this skill do?
Čech cohomology for local-to-global consistency verification in code structure and data schemas.
Where can I find the source code?
You can find the source code on GitHub using the link provided at the top of the page.
SKILL.md Source
# Sheaf Cohomology Skill: Local-to-Global Verification
**Status**: ✅ Production Ready
**Trit**: -1 (MINUS - validator/constraint)
**Color**: #2626D8 (Blue)
**Principle**: Local consistency → Global correctness
**Frame**: Čech cohomology with descent conditions
---
## Overview
**Sheaf Cohomology** validates that locally consistent data/code patches glue correctly into globally consistent structures. Uses:
1. **Čech cohomology**: H^n(U, F) obstruction classes
2. **Nerve of coverage**: N(U) simplicial complex from open cover
3. **Descent conditions**: Cocycle conditions for morphisms
4. **tree-sitter integration**: AST-level local consistency
**Correct by construction**: If local patches satisfy cocycle conditions, global structure is guaranteed.
## Core Formula
```
H⁰(U, F) = ker(d⁰) # Global sections (agree everywhere)
H¹(U, F) = ker(d¹)/im(d⁰) # Obstruction to gluing
H²(U, F) = ker(d²)/im(d¹) # Higher obstructions
```
For code verification:
```ruby
# Three patches (files/modules) are consistent iff:
# On U_ij ∩ U_jk ∩ U_ik: g_ij ∘ g_jk = g_ik (cocycle condition)
cocycle_satisfied?(patch_i, patch_j, patch_k)
== (compose(g_ij, g_jk) == g_ik)
```
## Why Sheaf Cohomology for Code?
1. **Module boundaries**: Each module is an "open set"
2. **Import/export**: Transition functions between patches
3. **Type consistency**: Cocycle = type compatibility
4. **Refactoring safety**: H¹ = 0 means safe global transform
## Gadgets
### 1. ČechCoverVerifier
Verify local consistency across code patches:
```ruby
verifier = SheafCohomology::CechCoverVerifier.new(
coverage: [:module_a, :module_b, :module_c]
)
verifier.add_transition(:module_a, :module_b, transition_ab)
verifier.add_transition(:module_b, :module_c, transition_bc)
verifier.add_transition(:module_a, :module_c, transition_ac)
verifier.cocycle_satisfied? # => true if g_ab ∘ g_bc = g_ac
verifier.h1_obstruction # => 0 if globally consistent
```
### 2. NerveConstructor
Build simplicial complex from coverage:
```ruby
nerve = SheafCohomology::NerveConstructor.new(
opens: file_modules,
intersections: shared_interfaces
)
nerve.simplices(0) # => vertices (modules)
nerve.simplices(1) # => edges (shared interfaces)
nerve.simplices(2) # => triangles (triple overlaps)
nerve.euler_characteristic # => χ(N(U))
```
### 3. DescentVerifier
Check morphism descent conditions:
```ruby
descent = SheafCohomology::DescentVerifier.new
descent.add_local_section(:patch_a, section_a)
descent.add_local_section(:patch_b, section_b)
descent.verify_descent! # => raises if descent fails
descent.global_section # => glued global section
```
### 4. TreeSitterSheaf
Integration with tree-sitter for AST verification:
```ruby
sheaf = SheafCohomology::TreeSitterSheaf.new(
language: :clojure,
coverage: :function_boundaries
)
sheaf.parse_file("src/core.clj")
sheaf.local_consistency_check # => per-function type consistency
sheaf.global_gluing_check # => cross-function compatibility
sheaf.h1_obstructions # => list of gluing failures
```
## Commands
```bash
# Verify sheaf consistency
just sheaf-check
# Check specific coverage
just sheaf-coverage src/
# Compute cohomology obstructions
just sheaf-h1
# Integration with tree-sitter
just sheaf-ast src/*.clj
```
## API
```ruby
require 'sheaf_cohomology'
# Create verifier
verifier = SheafCohomology::Verifier.new(
trit: -1,
coverage_strategy: :ast_boundaries
)
# Add patches from tree-sitter
verifier.add_patches_from_ast(parsed_files)
# Check consistency
result = verifier.verify!
result[:h0] # Global sections (fully consistent)
result[:h1] # Gluing obstructions
result[:h2] # Higher obstructions
result[:gf3_conserved] # GF(3) sum = 0 with triad
```
## Integration with GF(3) Triads
Forms valid triads with ERGODIC (0) and PLUS (+1) skills:
```
sheaf-cohomology (-1) ⊗ acsets (0) ⊗ gay-mcp (+1) = 0 ✓
sheaf-cohomology (-1) ⊗ unworld (0) ⊗ rama-gay-clojure (+1) = 0 ✓
sheaf-cohomology (-1) ⊗ glass-bead-game (0) ⊗ rubato-composer (+1) = 0 ✓
```
## Mathematical Foundation
### Čech Complex
```
C⁰ = ∏_i F(U_i)
C¹ = ∏_{i<j} F(U_i ∩ U_j)
C² = ∏_{i<j<k} F(U_i ∩ U_j ∩ U_k)
```
### Differential Maps
```
d⁰: C⁰ → C¹: (d⁰s)_{ij} = s_j|_{U_ij} - s_i|_{U_ij}
d¹: C¹ → C²: (d¹g)_{ijk} = g_jk - g_ik + g_ij
```
### Cocycle Condition
```
g ∈ Z¹(U, F) ⟺ d¹g = 0 ⟺ g_ij + g_jk = g_ik on triple overlaps
```
## Example Output
```
─── Sheaf Cohomology Verification ───
Coverage: 5 modules, 8 interfaces, 3 triple overlaps
Local Consistency:
✓ module_a ∩ module_b: types compatible
✓ module_b ∩ module_c: types compatible
✓ module_a ∩ module_c: types compatible
Čech Cohomology:
H⁰ = 1 (connected components)
H¹ = 0 (no gluing obstructions)
H² = 0 (no higher obstructions)
Global Structure: ✓ CONSISTENT
GF(3) Trit: -1 (MINUS/Validator)
```
---
**Skill Name**: sheaf-cohomology
**Type**: Local-to-Global Verification
**Trit**: -1 (MINUS)
**Color**: #2626D8 (Blue)
**GF(3)**: Forms valid triads with ERGODIC + PLUS skillsRelated Skills
Sheaf-Theoretic Coordination
**Category:** Phase 3 Core - Distributed Reasoning
sheaf-laplacian-coordination
Sheaf neural network coordination via graph Laplacians for distributed consensus and harmonic inference. Use when coordinating multi-agent systems, building sheaf-aware GNNs, or implementing distributed consensus protocols.
zx-calculus
Coecke's ZX-calculus for quantum circuit reasoning via string diagrams with Z-spiders (green) and X-spiders (red)
zulip-cogen
Zulip Cogen Skill 🐸⚡
zls-integration
zls-integration skill
zig
zig skill
zig-syrup-bci
Multimodal BCI pipeline in Zig: DSI-24 EEG, fNIRS mBLL, eye tracking IVT, LSL sync, EDF read/write, GF(3) conservation
zig-programming
zig-programming skill
zeroth-bot
Zeroth Bot - 3D-printed open-source humanoid robot platform for sim-to-real and RL research. Affordable entry point for humanoid robotics.
xlsx
Comprehensive spreadsheet creation, editing, and analysis with support
wycheproof
Google's Wycheproof test vectors for cryptographic implementation testing.
Writing Hookify Rules
This skill should be used when the user asks to "create a hookify rule", "write a hook rule", "configure hookify", "add a hookify rule", or needs guidance on hookify rule syntax and patterns.