---
description: "Project memory system for context recording and transfer using AI"
globs: ["**/.memory.json", "**/memory.json"]
imports:
- ".cursor/rules/core/memory_management.mdc"
commands:
CONTEXT:
description: "Manages AI context for the current session"
action: "manage_ai_context"
args:
- name: "operation"
type: "string"
description: "Operation to perform with the context"
enum: ["show", "update", "reset", "transfer"]
default: "show"
- name: "focus"
type: "string"
description: "Updates the current focus of the session"
required: false
- name: "preferences"
type: "array"
description: "Updates coding preferences"
items:
type: "string"
required: false
- name: "next_tasks"
type: "array"
description: "Updates the upcoming tasks to perform"
items:
type: "string"
required: false
- name: "to_file"
type: "string"
description: "Path where to save the extracted context"
required: false
SUMMARIZE:
description: "Generates a summary of the current conversation for the project memory"
action: "summarize_conversation"
args:
- name: "format"
type: "string"
description: "Summary format"
enum: ["concise", "detailed", "actionable"]
default: "concise"
- name: "focus"
type: "string"
description: "Aspect to focus on for the summary"
enum: ["decisions", "technical", "next_steps", "all"]
default: "all"
- name: "save_to_memory"
type: "boolean"
description: "Automatically save to project memory"
default: true
---
# Project Memory System with AI
The project memory system allows recording decisions, achievements, issues, and maintaining context between work sessions using AI.
## Objectives
- Maintain a historical record of development
- Preserve context between work sessions
- Facilitate knowledge transfer
- Assist in decision-making based on precedents
- Reduce ramp-up time when switching tasks
## Available Commands
### CONTEXT
This command allows managing the AI context for the current session.
**Examples:**
Show current context:
```
CONTEXT
```
Update session focus:
```
CONTEXT operation=update focus="Implementation of metadata analysis for SEO"
```
Update pending tasks:
```
CONTEXT operation=update next_tasks=["Complete US-04", "Refactor HeadingValidator class", "Implement integration tests"]
```
Transfer context to a file:
```
CONTEXT operation=transfer to_file="path/to/context.json"
```
Reset context:
```
CONTEXT operation=reset
```
### SUMMARIZE
This command generates a summary of the current conversation to store in the project memory.
**Examples:**
Generate concise summary:
```
SUMMARIZE
```
Generate detailed summary focused on technical decisions:
```
SUMMARIZE format=detailed focus=decisions
```
Generate actionable summary without saving it to memory:
```
SUMMARIZE format=actionable save_to_memory=false
```
### MEMORY
This command shows the contents of the development record.
**Examples:**
Show all memory:
```
MEMORY
```
Show only decisions:
```
MEMORY filter=decisions
```
Show a summary of AI context:
```
MEMORY filter=ai_context format=summary
```
### ACHIEVEMENT
This command records an achievement in the project memory.
**Examples:**
```
ACHIEVEMENT description="Completed refactoring of HeadingValidator class" category="development" impact_level="medium"
```
### ISSUE
This command records an issue in the project memory.
**Examples:**
```
ISSUE description="Detected inconsistency in the scoring algorithm" severity="medium" area="technical"
```
### DECISION
This command records a technical decision in the project memory.
**Examples:**
```
DECISION description="Adopt Strategy pattern for header validators" impact=["architecture", "development"] rationale="Improve extensibility and follow SOLID principles"
```
## Project Memory Workflow
1. **Session Start**: Use `CONTEXT` to load context from the previous session
2. **During Development**: Record achievements, issues, and decisions with the corresponding commands
3. **Session End**: Use `SUMMARIZE` to generate a summary of what was worked on
## Integration with Git Flow
The memory system automatically integrates with Git through hooks that:
1. Update memory with each commit
2. Extract decisions from commit messages
3. Detect technical debt patterns in the code
4. Update quality metrics when there are changes in tests
## Tips to Maximize Memory Value
- Record important decisions while working
- Use informative commit messages that start with types (feat, fix, refactor, etc.)
- Run `SUMMARIZE` at the end of significant sessions
- Update the current focus with `CONTEXT` when changing tasks
- Consult memory before making decisions similar to those already made
## Recommendations for Teams
- Periodically review recorded decisions
- Use memory for onboarding sessions
- Consult memory during planning
- Regularly update team coding preferences
## Learn More
For more information about memory management with AI, run:
```
HELP CONTEXT
```
