# CLAUDE.md - Project Context & Rules
## Memory Integration (AGIT)
You are connected to the `agit` MCP server, which acts as the long-term memory for this project.
### CRITICAL RULES
1. **Log User Intent:** When the user gives you a task, IMMEDIATELY call the tool `agit_log_step` with:
- `role`: "user"
- `category`: "intent"
- `content`: The user's request (summarized if very long)
2. **Log Your Plan:** BEFORE you write any code or edit files, call `agit_log_step` with:
- `role`: "ai"
- `category`: "reasoning"
- `content`: Your step-by-step plan for accomplishing the task
3. **Log Errors:** If you encounter an error or obstacle, call `agit_log_step` with:
- `role`: "ai"
- `category`: "error"
- `content`: Description of the error and your approach to fix it
4. **Read History:** If you are unsure of the current task or context:
- Call `agit_read_roadmap` to understand project goals
- Call `agit_get_recent_summaries` to see what was done recently
### Why This Matters
Every thought you log becomes part of the project's permanent history. When someone runs `agit log`, they will see not just WHAT changed, but WHY it changed and HOW the decision was made.
This creates a "Neural Graph" alongside the Git graph - code is the artifact, context is the source.