iterative-retrieval

Name: iterative-retrieval
Availability: InStock
Author: chomchomchomy

Automate iterative file and knowledge discovery with relevance scoring.

Updated Aug 23, 2026

Search Tags:#task-analysis #code-discovery #relevance-scoring #context-collection #iterative-search #multi-cycle

Ready to run

npx skills add https://github.com/chomchomchomy/antigravity-skills --skill iterative-retrieval-chomchomchomy

💬 Prompt for ChatGPT / Claude Chat▼

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

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill helps you identify the right files and background knowledge by performing up to three iterative search cycles to clarify complex tasks.

Core Features & Use Cases

Iterative search phases (DISPATCH, EVALUATE, REFINE, LOOP) to surface relevant sources.
Relevance scoring (0.0–1.0) to prioritize highly pertinent code and docs.
Flexible re-search: refine results with new keywords and re-evaluate candidates.

Quick Start

Initiate a three-cycle search to surface the most relevant files and context for your task.

Frequently Asked Questions about iterative-retrieval

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

FAQPage Schema

How do I find relevant files and context for complex debugging across multiple repositories?▼

To find relevant files and context for complex debugging, you can use an iterative search approach that runs up to three cycles of dispatching queries, evaluating relevance scores, and refining keywords to surface pertinent code across multi-repo projects.

What is iterative code discovery and how does relevance scoring work?▼

Iterative code discovery is a multi-cycle search process that evaluates and refines retrieved sources. Relevance scoring works by assigning a 0.0 to 1.0 priority value to candidates, helping you filter highly pertinent code and documentation from less relevant results.

How do I locate the right files for large-scale refactors when initial search results are insufficient?▼

To locate the right files for large-scale refactors, apply a looped refinement process that re-evaluates initial candidates and re-searches with new keywords, continuously refining context collection until the correct background knowledge is surfaced.

Can I automate background knowledge collection for feature development in multi-repo projects?▼

You can automate background knowledge collection for feature development in multi-repo projects by initiating a structured search cycle that dispatches queries, evaluates relevance, and loops through refinements to clarify complex task contexts.

What's the best way to surface relevant code for a complex task when a single search is not enough?▼

The best way to surface relevant code when a single search is insufficient is an iterative retrieval method, which executes up to three search phases that evaluate and refine results based on dynamic relevance scoring and contextual gaps.

Does iterative file discovery work for tasks that require deep context across different repositories?▼

Iterative file discovery works for deep context tasks by applying up to three search cycles across multi-repo projects, utilizing dispatch, evaluate, refine, and loop phases to thoroughly clarify complex task requirements and relevant background knowledge.