EXAMPLES.md•1.69 kB
# Sequential Thinking MCP Server Test
## Test Scenarios
### Scenario 1: Basic Sequential Thinking
```json
{
"tool": "sequential_thinking",
"arguments": {
"thought": "Let me break down this complex math problem step by step. First, I need to identify what we're solving for.",
"nextThoughtNeeded": true,
"thoughtNumber": 1,
"totalThoughts": 3
}
}
```
### Scenario 2: Revision of Previous Thought
```json
{
"tool": "sequential_thinking",
"arguments": {
"thought": "Actually, I realize my previous approach was flawed. Let me reconsider the problem from a different angle.",
"nextThoughtNeeded": true,
"thoughtNumber": 3,
"totalThoughts": 4,
"isRevision": true,
"revisesThought": 2
}
}
```
### Scenario 3: Branching Reasoning
```json
{
"tool": "sequential_thinking",
"arguments": {
"thought": "This opens up two possible approaches. Let me explore the algorithmic solution first.",
"nextThoughtNeeded": true,
"thoughtNumber": 4,
"totalThoughts": 6,
"branchFromThought": 2,
"branchId": "algorithmic"
}
}
```
### Scenario 4: Getting Summary
```json
{
"tool": "get_thinking_summary",
"arguments": {}
}
```
### Scenario 5: Reset State
```json
{
"tool": "reset_thinking",
"arguments": {}
}
```
## Usage Instructions
1. Build the server: `npm run build`
2. Run the server: `npm run dev`
3. Configure in Claude Desktop or VS Code MCP settings
4. Use the tools to engage in structured thinking processes
## Expected Outputs
The server provides formatted responses showing:
- Current thought number and progress
- Revision and branching indicators
- Completion status
- Rich formatting for readability