23blocks-os-ai-maestro-documentation-search_skill

Overview

This skill proactively searches auto-generated project documentation whenever it receives any user instruction. It prioritizes function signatures, class docs, API descriptions, and code comments so implementations follow existing patterns and avoid errors. The goal is to ensure correct usage of existing code before writing or modifying anything.

How this skill works

On every user instruction the skill immediately runs semantic and keyword documentation searches to find relevant functions, classes, modules, and guides. It inspects signatures, parameters, return types, and author comments, then reports findings before any code changes or calls are made. If tools are missing or docs are empty, it surfaces clear next steps (indexing or installing helpers).

When to use it

• Before implementing any new feature or service
• When calling or modifying a function or method
• When integrating with an API or endpoint
• When fixing a bug or refactoring code
• When onboarding to an unfamiliar codebase

Best practices

• Always run docs-search first—treat it as the default reaction to user instructions
• Search by both semantic query and exact keyword for function/class names
• Combine memory and code-graph searches with docs-search for full context
• Index documentation regularly so searches return useful results
• Report no-docs findings immediately and offer to index or inspect code directly

Example use cases

• User asks to implement authentication — run semantic and keyword searches for auth-related docs before coding
• User requests modifying PaymentService — find class docs and method signatures prior to changes
• User mentions a function name — run keyword search to confirm exact parameters and return values
• On a bug report referencing a method Z — pull the doc for Z and any related guide sections first
• Before calling an internal API endpoint — verify docs for expected payloads and error responses