Is it safe for large datasets?

Yes, it implements guardrails to prevent pasting large raw chunks into the chat, instead managing them via local state files under the .claude directory.

Does this skill require external tools?

Yes, it requires a local Python 3 environment to run the persistent REPL scripts that manage context and chunking.

Can I define how the data is chunked?

Yes, you can specify chunk size and overlap, and the skill prefers semantic chunking for formats like JSON, Markdown, or timestamped logs.

What is the primary benefit of the RLM skill?

It allows Claude to process files that are too large for the standard chat context window by breaking them into manageable chunks for iterative analysis.

How does the recursive sub-agent system work?

The main Claude agent orchestrates the workflow, while the rlm-subcall agent performs specific analysis on individual data chunks to keep the main conversation lean.

RLM Recursive Context Processor

Name: RLM Recursive Context Processor
Author: Bogyie

byBogyie

0•

데이터 과학 및 ML

Processes and analyzes massive files through a recursive language model loop and persistent local Python environment.

The RLM skill enables Claude to handle long-context tasks that exceed standard memory limits, such as analyzing massive logs, transcripts, or scraped datasets. By utilizing a persistent local Python REPL and a specialized sub-agent (rlm-subcall), the skill scouts, chunks, and extracts information iteratively. This recursive approach allows for deep, granular analysis of large-scale data while maintaining a structured synthesis in the main conversation, preventing context overflow and ensuring precision.

주요 기능

01Smart 'scouting' to preview large files without loading them

02Persistent local Python REPL for stateful data manipulation

030 GitHub stars

04Customizable semantic or character-based chunking strategies

05Recursive sub-agent delegation for parallel chunk analysis

06Automated synthesis of findings from multiple data segments

사용 사례

01Analyzing gigabytes of server logs to identify complex error patterns

02Querying and extracting specific data from massive technical documentation sets

03Synthesizing insights from large-scale web scraping outputs or long transcripts

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add bogyie/bogyie-plugins skills

For use in Claude.ai and ChatGPT

Download Skill