Key Findings

Discovered Topics

📝
Agent & Task
Code style rules, event handling, and error patterns
agent
0.5751
task
0.499
skill
0.4584
change
0.4084
read
0.3833
rule
0.3365
git
0.3119
dont
0.2768
tool
0.2726
commit
0.2696

This topic captures how developers structure their requests to Claude, covering task workflows, documentation patterns, and user context. It reflects the meta-layer of how teams organize their Claude interactions beyond just code generation.

📊
Agentsmd & Instruction
Cross-cutting concerns: validation, APIs, and error handling
agentsmd
2.052
instruction
0.1613
read
0.1258
agent
0.0815
follow
0.0719
guideline
0.0601
repo
0.0563
convention
0.0436
canonical
0.0413
making
0.0379

This topic represents a cluster of related terms discovered through topic modeling analysis. The word weights indicate how strongly each term is associated with this theme across the analyzed repositories.

⚛️
Frontend Development
React, TypeScript, npm tooling, and client-side patterns
npm
0.7103
component
0.4955
pnpm
0.443
react
0.3465
typescript
0.3047
api
0.2902
app
0.2747
backend
0.2688
server
0.2653
frontend
0.2586

Frontend-focused CLAUDE.md files are among the most detailed. The prevalence of "npm," "typescript," and "react" reflects the dominance of the React/TypeScript ecosystem in projects using Claude Code. These files typically specify component patterns, state management, and build tooling.

🔧
Data & APIs
API design, databases, server config, and integrations
python
0.5079
model
0.3271
data
0.2814
make
0.2405
config
0.2334
image
0.2203
script
0.1897
cli
0.1874
install
0.1829
docker
0.1739

This topic represents a cluster of related terms discovered through topic modeling analysis. The word weights indicate how strongly each term is associated with this theme across the analyzed repositories.

Cargo & Crate
Documentation structure, user context, and task workflows
cargo
1.1954
crate
0.4513
rust
0.2619
workspace
0.1892
cli
0.1642
bun
0.1505
make
0.1499
type
0.1284
format
0.1263
release
0.115

This topic represents a cluster of related terms discovered through topic modeling analysis. The word weights indicate how strongly each term is associated with this theme across the analyzed repositories.