Find your way around an unfamiliar repository

Start with the repository structure

ashlr__tree accepts a depth limit and a regex pattern filter. In a Git repository it uses git ls-files; outside Git it walks directories with an exclusion set. Use that structural view to choose a narrower area to investigate.

Sources: Tree implementation and registered filters

Ask a concrete architecture question

ashlr__orient combines a tree scan, file discovery, compact reads and local model synthesis. Its registered inputs are query, dir and a depth of quick or thorough. Use the discovered paths as leads to inspect, not as a substitute for reading the code.

Sources: Registered orientation workflow and inputs

Verify before changing behavior

Orientation uses compacted file reads, which can omit the middle of a file. Follow the discovered paths and use your host's native range reader to inspect any omitted code before deciding where a fix belongs.

Sources: Orientation discovery; What compact reads preserve

Try this workflow

1. Install and verify the plugin for your coding host.
2. Ask for a shallow tree and identify the directories relevant to the report.
3. Ask one architecture question, inspect the cited paths, and write a short change plan with tests.

Use ashlr__tree to map this project, then ask ashlr__orient how an incoming request reaches persistence. Show the discovered paths and identify which ones need a closer read.

Install and verify your host integration or open the quick install commands.