When is this workflow automatically triggered?

The workflow is auto-loaded whenever you ask Claude questions regarding code architecture, patterns, project decisions, or specific technical choices.

How does the memory-first workflow save tokens?

It directs Claude to query a lightweight SQLite database for specific facts and past decisions rather than re-reading and processing large source files for every technical question.

Can I manually invoke this workflow?

While it is designed to run automatically behind the scenes, you can trigger specific memory actions using commands like memory.recall, memory.decisions, or memory.conventions.

Does this skill prevent Claude from reading my code?

No, it simply changes the order of operations. Claude will still explore files using Read or Grep if the memory is insufficient or if you request specific implementation details.

Memory-First Research Workflow

Name: Memory-First Research Workflow
Author: codenamev

bycodenamev

•

Productivity & Workflow

Prioritizes long-term project memory and architectural history over redundant file exploration to optimize research efficiency.

About

The memory-first-workflow is a strategic research pattern designed to make Claude more efficient by consulting persistent project memory before scanning the codebase. By utilizing SQLite-backed MCP tools to recall past decisions, architectural patterns, and coding conventions, this skill significantly reduces token consumption and prevents redundant processing of large files. It ensures that Claude's responses remain contextually aligned with established project history while providing a structured path for investigating missing information only when necessary.

Key Features

Integrated conflict detection to surface architectural disputes early
Token-efficient research patterns that minimize unnecessary file reads
Automated memory-first lookup for technical and architectural queries
Source-aware citation system distinguishing between recalled and discovered data
8 GitHub stars
Specialized shortcuts for retrieving decisions, architecture, and conventions

Use Cases

Onboarding to a large codebase by querying established coding conventions
Consulting architectural history before implementing major system changes
Reducing API costs by leveraging local SQLite memory for common questions

About

Key Features

Integrated conflict detection to surface architectural disputes early
Token-efficient research patterns that minimize unnecessary file reads
Automated memory-first lookup for technical and architectural queries
Source-aware citation system distinguishing between recalled and discovered data
8 GitHub stars
Specialized shortcuts for retrieving decisions, architecture, and conventions

Use Cases

Onboarding to a large codebase by querying established coding conventions
Consulting architectural history before implementing major system changes
Reducing API costs by leveraging local SQLite memory for common questions