Agent Guidelines for LangGraphGo

Build/Lint/Test Commands

make test

- Run all tests; for single test:

go test -v ./path/to/pkg -run TestName

Use
```
make fmt
```
to format code with gofmt before committing
Exported types/functions: PascalCase; private: camelCase
Define errors as package-level variables:
```
var (ErrX = errors.New("..."))
```
Write godoc comments for all exported identifiers
Tests: Use table-driven tests with
```
t.Run()
```
,
```
t.Parallel()
```
where appropriate
Always check errors (linted by golangci-lint)
Run
```
make check
```
(fmt-check, vet, lint) and
```
make test
```
before committing

graph: Core graph construction and execution engine with state management
memory: Various memory strategies for conversational AI applications
prebuilt: Ready-to-use agent implementations (ReAct, Supervisor, Planning, etc.)
ptc: Programmatic tool calling using generated code (Python, JavaScript, Shell)
rag: Retrieval-Augmented Generation with vector and GraphRAG support
store: Checkpoint storage backends (SQLite, PostgreSQL, Redis)
tool: Collection of tools for web search, file ops, code execution
log: Simple leveled logging interface for applications
adapter: Integration adapters for GoSkills, MCP, and external systems

make test

- Run all tests; for single test:

go test -v ./path/to/pkg -run TestName

Use
```
make fmt
```
to format code with gofmt before committing
Exported types/functions: PascalCase; private: camelCase
Define errors as package-level variables:
```
var (ErrX = errors.New("..."))
```
Write godoc comments for all exported identifiers
Tests: Use table-driven tests with
```
t.Run()
```
,
```
t.Parallel()
```
where appropriate
Always check errors (linted by golangci-lint)
Run
```
make check
```
(fmt-check, vet, lint) and
```
make test
```
before committing

graph: Core graph construction and execution engine with state management
memory: Various memory strategies for conversational AI applications
prebuilt: Ready-to-use agent implementations (ReAct, Supervisor, Planning, etc.)
ptc: Programmatic tool calling using generated code (Python, JavaScript, Shell)
rag: Retrieval-Augmented Generation with vector and GraphRAG support
store: Checkpoint storage backends (SQLite, PostgreSQL, Redis)
tool: Collection of tools for web search, file ops, code execution
log: Simple leveled logging interface for applications
adapter: Integration adapters for GoSkills, MCP, and external systems