scan

Analyze a codebase with four parallel mapper agents into .planning/codebase.

Updated Aug 23, 2026

Search Tags:#automation #architecture #codebase #analysis #reconnaissance #mapping #parallelism

Ready to run

npx skills add https://github.com/SienkLogic/towline --skill scan-sienklogic

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: scan
Source: https://github.com/SienkLogic/towline/tree/main/plugins/dev/skills/scan
Command: npx skills add https://github.com/SienkLogic/towline --skill scan-sienklogic

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Analyze an existing codebase to produce a comprehensive map of its structure, architecture, conventions, and concerns for brownfield projects.

Core Features & Use Cases

Spawn four parallel Task(subagent_type: "dev:towline-codebase-mapper") agents for analysis and mapping.
Delegate reading and pattern recognition to specialized mapper subagents to maintain context isolation.
Generate a coherent reconnaissance and architecture outline that informs planning and implementation.

Quick Start

Run the scan skill to generate a full codebase map and begin analysis with four parallel mapper agents.

Frequently Asked Questions about scan

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

FAQPage Schema

How do I map and understand an existing codebase structure?▼

Map an existing codebase by running an automated reconnaissance analysis that spawns four parallel mapper agents to evaluate architecture, conventions, and concerns, generating a comprehensive structural outline.

What is the best way to analyze brownfield project architecture automatically?▼

Automated brownfield project architecture analysis is best handled by delegating reading and pattern recognition to specialized parallel mapper subagents, maintaining context isolation while producing a coherent architecture outline.

How does parallel codebase mapping work for large projects?▼

Parallel codebase mapping works by coordinating four parallel Task agents acting as specialized mappers, independently analyzing different codebase facets while maintaining context isolation to generate a multi-faceted structural map.

Can I use automated codebase reconnaissance without modifying my source files?▼

Automated codebase reconnaissance can be used without modifying source files because the analysis enforces a read-only, agent-driven workflow, safely mapping architecture while writing outputs only to the .planning/codebase directory.

Where are the architecture analysis results stored during codebase scanning?▼

Codebase scanning architecture analysis results are stored as defined outputs in the .planning/codebase directory, generating concise frontmatter-focused results to inform downstream planning and implementation tasks.