iterative-retrieval

Iteratively search, evaluate, and refine file retrieval for subagent context.

1|Updated Mar 4, 2026

Search Tags:#context retrieval #information retrieval #rag #multi-agent systems #agentic workflow #codebase exploration

Ready to run

npx skills add https://github.com/flatrick/mdt --skill iterative-retrieval-flatrick

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: iterative-retrieval
Source: https://github.com/flatrick/mdt/tree/main/skills/iterative-retrieval
Command: npx skills add https://github.com/flatrick/mdt --skill iterative-retrieval-flatrick

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill addresses the challenge where subagents lack the precise context needed to perform their tasks, preventing them from knowing which files or information are relevant upfront.

Core Features & Use Cases

Progressive Context Refinement: Iteratively searches and evaluates files to build the optimal context for subagents.
Handles Unpredictable Needs: Ideal for multi-agent workflows where context requirements evolve during execution.
RAG Optimization: Improves Retrieval Augmented Generation pipelines by focusing on relevant code.
Use Case: When debugging a complex authentication issue, this skill can intelligently discover and retrieve only the most relevant authentication, session, and token-related files, avoiding overwhelming the agent with the entire codebase.

Quick Start

Use the iterative-retrieval skill to find files related to 'user authentication' in the codebase.

Frequently Asked Questions about iterative-retrieval

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I improve context retrieval for subagents in a multi-agent workflow?▼

Improving context retrieval for subagents requires an iterative process that dynamically searches, evaluates, and refines files based on task relevance. This approach progressively builds optimal context without upfront knowledge of relevant code.

What is iterative retrieval in RAG pipelines for codebase exploration?▼

Iterative retrieval in RAG pipelines is a pattern that progressively refines context by searching and evaluating files to identify gaps. It optimizes token usage by focusing exclusively on high-relevance code for the specific task.

How do I stop overwhelming my agent with an entire codebase during complex debugging?▼

To avoid overwhelming an agent during complex debugging, implement an iterative retrieval pattern that intelligently discovers and retrieves only the most relevant files. It dynamically filters information based on identified task relevance and context gaps.

Can I use progressive context refinement for unpredictable agentic workflows?▼

Yes, progressive context refinement is specifically designed for unpredictable agentic workflows. It handles evolving context requirements by dynamically searching and evaluating file retrieval during execution to supply precise information.

Does iterative file retrieval work without knowing the relevant files upfront?▼

Yes, iterative file retrieval works precisely because it solves the problem of not knowing relevant files upfront. It dynamically evaluates and refines context retrieval based on task relevance and identified information gaps.