When should I use the recall skill vs a standard query?

Use recall when you have 'why' or 'how' questions about existing structure and decisions. Use standard queries when you want to modify code or perform new research.

How does it know where my documentation is stored?

It checks your .smith.local.yaml file for a custom output_directory value, otherwise it defaults to the docs/ folder in your project root.

What documents does Smith Recall search?

It primarily searches markdown files in the research and plans subdirectories of your project's defined documentation folder.

Does it only search existing documents?

No. If existing research or plans do not fully answer your question, the skill automatically falls back to standard codebase exploration using tools like Grep and Glob.

Smith Recall

Name: Smith Recall
Author: brettbuddin

bybrettbuddin

•

Learning & Documentation

Retrieves project architecture, design rationale, and historical context from existing Smith research and planning documents.

Smith Recall is a context-aware lookup skill for Claude Code that prioritizes existing project intelligence over redundant codebase scanning. It automatically locates research papers and implementation plans within a project's documentation directory to answer complex questions about architecture, design decisions, and system behavior. By referencing previous Smith commands and historical documents, it ensures Claude provides answers consistent with established project conventions and past research, falling back to direct code exploration only when necessary.

Key Features

01Citation-aware responses that reference specific source file paths

02Configurable path resolution via .smith.local.yaml settings

03Automated document discovery in Smith research and plan directories

04Smart fallback to codebase exploration when documentation is incomplete

053 GitHub stars

06Contextual prioritization of recently modified or session-relevant files

Use Cases

01Recalling why specific design trade-offs were made during previous research phases

02Understanding high-level architecture without scanning every source file

03Onboarding to complex projects by querying existing implementation plans

Key Features

01Citation-aware responses that reference specific source file paths

02Configurable path resolution via .smith.local.yaml settings

03Automated document discovery in Smith research and plan directories

04Smart fallback to codebase exploration when documentation is incomplete

053 GitHub stars

06Contextual prioritization of recently modified or session-relevant files

Use Cases

01Recalling why specific design trade-offs were made during previous research phases

02Understanding high-level architecture without scanning every source file

03Onboarding to complex projects by querying existing implementation plans