Is this skill only compatible with Nuxt 4 projects?

While the cellm plugin is optimized for the Nuxt 4 stack, the indexing skill is designed with universal scope and can be used to manage conventions for any codebase.

What happens when I delete code that was previously indexed?

The skill includes a cleanup routine that identifies atoms whose source evidence no longer exists in the project and automatically archives them to prevent stale context.

How does the skill prevent duplicate entries in the knowledge base?

Before adding any new entry, the skill performs a mandatory search of existing knowledge atoms to ensure the information isn't already indexed.

Does this skill create index files in my project directory?

No, the Cellm indexer persists all data as knowledge atoms within the Oracle system and is strictly prohibited from writing index files to your local disk.

Why are the index entries limited to one line?

Keeping entries short and factual ensures that the AI can process the context efficiently without being overwhelmed by verbose or conversational descriptions.

Cellm Context Indexer

Name: Cellm Context Indexer
Author: murillodutt

bymurillodutt

0•

生産性とワークフロー

Maintains a persistent project convention index using Oracle knowledge atoms to optimize AI context engineering.

Cellm Context Indexer is a specialized skill designed to synchronize project-specific patterns and conventions with an external Oracle knowledge base. By scanning project files and extracting factual, evidence-backed insights, it ensures that Claude has access to an up-to-date 'source of truth' for the codebase without cluttering the local disk with temporary index files. It is particularly effective for Nuxt 4 projects, providing a streamlined way to manage project standards through deduplicated knowledge atoms and automatic cleanup of stale data.

主な機能

010 GitHub stars

02Optimized context engineering specifically for Nuxt 4 stack conventions

03Automatic archival of stale atoms when source evidence is removed

04Automated project pattern discovery via advanced Grep and Glob searches

05Intelligent deduplication logic to prevent redundant knowledge entries

06Disk-free persistence using Oracle knowledge atoms for cleaner repositories

ユースケース

01Pruning outdated project documentation atoms to maintain an accurate AI knowledge base

02Establishing a shared pattern index for a new Nuxt 4 codebase to assist AI onboarding

03Synchronizing project conventions after a major architectural refactor

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add murillodutt/cellm index

For use in Claude.ai and ChatGPT