A Claude Code plugin that automatically creates and maintains THINKING_LOG.md files documenting Claude's reasoning process, decisions, and problem-solving approach during implementation tasks.
- Structured Documentation: Organized entries with Flow, Thinking, Decision, Action, Result format
- Visual Diagrams: Mermaid diagrams for flowcharts, sequence diagrams, mind maps, and git graphs
- Privacy-Aware: Default storage outside project source to avoid committing AI files
- Slash Command:
/thinking-logfor quick access to common operations - Auto-Trigger: Skill activates automatically during complex implementations
/plugin https://github.com/mduongvandinh/claude-thinking-log-plugin- Clone or download this repository
- Copy to your Claude Code plugins directory:
cp -r thinking-log-plugin ~/.claude/plugins/
# Initialize a new thinking log
/thinking-log init
# Add an entry
/thinking-log add --title "Implemented authentication"
# Add session summary
/thinking-log summary
# Show log status
/thinking-log showThe skill automatically activates when:
- User requests documentation of thinking/reasoning process
- Working on complex multi-step implementations
- User mentions "thinking log", "reasoning log", "decision log"
- Creating audit trails for code changes
thinking-log-plugin/
├── .claude-plugin/
│ └── plugin.json # Plugin manifest
├── commands/
│ └── thinking-log.md # /thinking-log command
├── skills/
│ └── thinking-log/
│ ├── SKILL.md # Main skill definition
│ └── references/
│ ├── advanced-patterns.md # Complex diagram patterns
│ └── starter-template.md # New log template
└── README.md
# Thinking Log - {Project Name}
## Session: {DATE}
### {Timestamp} - {Task Title}
**Flow:**
[Mermaid diagram]
**Thinking:**
[Raw reasoning]
**Decision:** [What was decided]
**Action:** [What was done]
**Result:** [Outcome]| Type | Use Case |
|---|---|
| Analysis | Understanding code/requirements |
| Decision | Choosing between alternatives |
| Implementation | Tracking build steps |
| Error Resolution | Documenting troubleshooting |
| Branch/Merge | Git operations |
| Diagram | Syntax | Best For |
|---|---|---|
| Flowchart | flowchart LR/TD |
Process flows |
| Sequence | sequenceDiagram |
API interactions |
| Mind Map | mindmap |
Concepts/architecture |
| Git Graph | gitGraph |
Branch strategy |
Default (recommended):
~/.claude/projects/{project-name}/docs/THINKING_LOG.md
Project-local (with --project flag):
{project-root}/docs/THINKING_LOG.md
The default location keeps AI-related files outside project source, useful when:
- Project policies prohibit AI-generated files
- Multiple developers share the project
- You want thinking logs private
Contributions are welcome! Please:
- Fork the repository
- Create a feature branch
- Submit a pull request
MIT License