Can Oracle help reduce my token usage?

Yes, Oracle uses a tiered context loading strategy to inject only the most relevant high-priority information, preventing context window bloat and saving tokens.

Does Oracle require manual data entry?

While you can manually query and update it, Oracle is designed to automatically monitor interactions and update the knowledge base at the end of each session.

How does Oracle store project knowledge?

Oracle stores all project knowledge locally in a `.oracle/` directory using human-readable JSON and Markdown files, ensuring your data stays within your repository.

How does Oracle learn from its mistakes?

When you provide a correction, Oracle records what was wrong and what the correct approach is, flagging it as a high-priority item to be recalled in similar future contexts.

Oracle Project Memory

Name: Oracle Project Memory
Author: Overlord-Z

byOverlord-Z

생산성 및 워크플로

Maintains project-specific institutional knowledge and learns from user feedback to optimize context and prevent repetitive errors.

소개

Oracle is a sophisticated memory and learning system for Claude Code that builds a persistent knowledge base for your software projects. By tracking interactions, recording corrections, and identifying recurring patterns, it ensures that architectural decisions and preferred conventions are maintained across multiple coding sessions. It intelligently manages token usage through tiered context injection, surfacing only the most relevant patterns and 'gotchas' for your current task while automating repetitive workflows through detected patterns.

주요 기능

Strategic Context Injection: Implements tiered loading to optimize token usage and keep context relevant.
Persistent Knowledge Base: Stores patterns, preferences, and pitfalls in a structured, searchable format.
Session Recording: Tracks decisions, successes, and failures in a chronological project timeline.
Automation Discovery: Identifies repetitive tasks and generates helper scripts to improve efficiency.
Correction Learning: Automatically updates behavior and flags mistakes to avoid repeating them in the future.
0 GitHub stars

사용 사례

Maintaining consistent architectural patterns and style preferences across long-running development projects.
Preventing the AI from repeating specific mistakes or using deprecated libraries after being corrected once.
Providing instant project onboarding by surfacing institutional knowledge and historical decisions to the LLM.

소개

주요 기능

Strategic Context Injection: Implements tiered loading to optimize token usage and keep context relevant.
Persistent Knowledge Base: Stores patterns, preferences, and pitfalls in a structured, searchable format.
Session Recording: Tracks decisions, successes, and failures in a chronological project timeline.
Automation Discovery: Identifies repetitive tasks and generates helper scripts to improve efficiency.
Correction Learning: Automatically updates behavior and flags mistakes to avoid repeating them in the future.
0 GitHub stars

사용 사례

Maintaining consistent architectural patterns and style preferences across long-running development projects.
Preventing the AI from repeating specific mistakes or using deprecated libraries after being corrected once.
Providing instant project onboarding by surfacing institutional knowledge and historical decisions to the LLM.