name	memory-augmented-dev
description	Development with persistent memory checks and automatic logging

Memory-Augmented Development

Purpose

Ensures every development task leverages past learnings and contributes to organizational memory. Search before implementing, document after completing.

When This Activates

User asks to implement a feature
User requests bug fixes or refactoring
Any coding task that could benefit from past patterns
Keywords: "implement", "build", "fix", "refactor", "add feature"

Instructions

Phase 1: Research (Before Implementation)

Semantic Search for Patterns Use search-enhanced to find relevant past work:

search-enhanced(
  query: "<feature area> implementation patterns",
  filters: {type: "code"}
)

Review Past Decisions

list-memories-by-tag(["<feature-area>", "architecture", "decisions"])

Check Recent Context

get-recent-context(project: "<current-project>")

Analyze Retrieved Memories
- What patterns were successful?
- What mistakes were made?
- What decisions inform this work?

Phase 2: Implementation

Apply Patterns Found
- Use established code patterns
- Follow past architectural decisions
- Avoid documented mistakes
Note Deviations
- If deviating from patterns, document why
- Prepare justification for memory storage

Phase 3: Documentation (After Implementation)

Store Memory with Rich Metadata

store-dev-memory({
  type: "code",
  content: "Detailed description of implementation",
  project: "<project-name>",
  tags: ["<feature>", "<technology>", "<pattern-used>"],
  metadata: {
    implementation_status: "complete",
    key_decisions: ["Decision 1", "Decision 2"],
    files_created: ["file1.py", "file2.py"],
    files_modified: ["existing.py"],
    code_changes: "Summary of major changes",
    dependencies_added: ["package1", "package2"],
    testing_notes: "How to test this"
  },
  relationships: [
    {
      memory_id: "<related-memory-hash>",
      type: "builds_on"
    }
  ]
})

Store Decisions Separately

store-dev-memory({
  type: "decision",
  content: "Why we chose approach X over Y",
  project: "<project-name>",
  tags: ["decision", "<topic>"],
  metadata: {
    alternatives_considered: ["Approach Y", "Approach Z"],
    decision_rationale: "Explanation",
    decision_date: "<date>",
    who_decided: "<name>"
  }
})

Tool Reference

search-enhanced

Purpose: Semantic search across all memories
Parameters:
- query (string): Search terms
- filters (object): Optional filters
  - type: "code" | "decision" | "conversation" | "reference"
  - tags: Array of tags to match
  - project: Project name
  - date_from, date_to: Date range
- limit (int): Max results (default: 10)
Returns: Array of memories with similarity scores
Example: search-enhanced("authentication JWT patterns", filters={type: "code"})

list-memories-by-tag

Purpose: Get all memories with specific tags
Parameters:
- tags (array): Tags to search for
- project (string): Optional project filter
Returns: Array of memories
Example: list-memories-by-tag(["authentication", "security"])

get-recent-context

Purpose: Get recent memories for session continuity
Parameters:
- project (string): Project name
- limit (int): Number of recent items (default: 10)
Returns: Array of recent memories, ordered by date
Example: get-recent-context(project: "hrdag-api")

store-dev-memory

Purpose: Store new memory with full metadata
Parameters: See Phase 3 examples above
Returns: Memory ID (hash)
Example: See detailed examples below

get-dev-memory

Purpose: Retrieve specific memory by ID
Parameters:
- memory_id (string): Memory hash ID
Returns: Full memory object
Example: get-dev-memory(memory_id: "abc123def456")

Examples

Example 1: Implementing Authentication

User says: "Add JWT authentication to the API"

You do:

Research

search-enhanced("authentication JWT patterns", filters={type: "code"})
list-memories-by-tag(["authentication", "security"])

Result: Found memory showing JWT implementation from 3 months ago

Implement Apply the JWT pattern found, using similar code structure

Document

store-dev-memory({
  type: "code",
  content: "Implemented JWT authentication with refresh tokens for API v2",
  project: "hrdag-api",
  tags: ["authentication", "jwt", "security", "api-v2"],
  metadata: {
    implementation_status: "complete",
    key_decisions: [
      "Used JWT over session-based auth for stateless API",
      "Refresh token rotation every 7 days"
    ],
    files_created: ["auth/jwt_manager.py", "auth/middleware.py"],
    files_modified: ["api/routes.py", "config/settings.py"],
    dependencies_added: ["PyJWT==2.8.0"],
    testing_notes: "Test with curl scripts in tests/auth/"
  },
  relationships: [{
    memory_id: "<hash-of-3-month-old-auth-memory>",
    type: "builds_on"
  }]
})

Example 2: Fixing Data Recovery Bug

User says: "The zip disk recovery keeps failing on bad sectors"