skills.rest
andrew-starosciak
andrew-starosciakCOMMUNITY
Explore all skills by this author →
𝕏@andrewscodingGitHub

iterative-retrieval

Refine subagent context through a four-stage iterative retrieval loop.

2|1|Updated Oct 4, 2025
Search Tags:#workflow#multi-agent#information-gathering#context-management#retrieval-augmented#iterative-retrieval#subagent-context
One-click install
npx skills add https://github.com/andrew-starosciak/deep-algo --skill iterative-retrieval-andrew-starosciak
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: iterative-retrieval
Source: https://github.com/andrew-starosciak/deep-algo/tree/main/.claude/docs/zh-TW/skills/iterative-retrieval
Command: npx skills add https://github.com/andrew-starosciak/deep-algo --skill iterative-retrieval-andrew-starosciak

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill provides a structured pattern to progressively refine the information context used by subagents, preventing context overload while ensuring critical information is surfaced.

Core Features & Use Cases

  • Iterative context refinement: A four-stage loop (DISPATCH, EVALUATE, REFINE, LOOP) guides context gathering and narrowing.
  • Relevance-driven discovery: Systematically scores candidate files and identifies essential missing context.
  • Controlled growth: Limits the number of cycles to three to avoid overfitting or drift.
  • Agent workflow optimization: Improves multi-agent coordination by surfacing coherent, task-focused context.

Quick Start

Start with a broad repository query for potential files, evaluate relevance, refine your search terms, and loop up to three cycles to assemble the required context.

Frequently Asked Questions about iterative-retrieval

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

FAQPage Schema
How do I refine context for subagents without overloading them with irrelevant files?▼

Iterative context refinement uses a four-stage loop to progressively narrow broad repository queries, scoring candidate files for relevance so subagents receive concise, task-focused context without overload.

What is the best way to gather missing context in a multi-agent workflow?▼

Multi-agent context gathering works best with an iterative retrieval loop that evaluates dispatched results, refines search terms, and identifies missing context through up to three structured cycles.

How does iterative retrieval improve subagent context management?▼

Iterative retrieval improves subagent context management by applying relevance scoring and query refinement across repeated cycles, ensuring that only high-relevance files are assembled into actionable context.

Can I limit context refinement cycles to prevent search drift in multi-agent workflows?▼

Yes, iterative retrieval limits the refinement cycle to a maximum of three loops, preventing overfitting or drift while systematically surfacing essential missing context for subagents.

When do I need iterative retrieval for assembling agent context?▼

You need iterative retrieval when subagents start with broad queries and must progressively narrow results to identify high-relevance files and missing context without exceeding three cycles.

What are the four stages of the iterative retrieval loop for context management?▼

The iterative retrieval loop consists of DISPATCH, EVALUATE, REFINE, and LOOP stages, guiding context gathering by dispatching queries, evaluating relevance, refining search terms, and cycling up to three times.

Related Skills

Explore More Software Engineering Skills→
AtlasRoXAtlasRoX

iterative-retrieval

Iteratively refine search queries to retrieve relevant code context for AI agents.

Community
Advanced
BuildSmarterAIBuildSmarterAI

iterative-retrieval

Automates progressive codebase context collection for multi-agent workflows via iterative refinement cycles.

Community
Advanced
linnefromicelinnefromice

iterative-retrieval

Refine retrieved context across iterative cycles for subagent workflows.

Community
Advanced
flag3flag3
1

iterative-retrieval

Guide subagents through a 4-phase retrieval loop to refine information.

Community
Basic