<2-3 sentences summarising what the project addresses and solves, the target audience, and the benefit.> konbakuyomu/smartsearch provides a CLI-first, skill-driven web research framework designed for AI agents and terminal users. It offers a reproducible command layer for live search, source discovery, page fetching, site mapping, provider diagnostics, and offline Deep Research planning, enabling efficient and structured information retrieval. The tool distinguishes between fast, live search results and detailed, offline research planning, catering to diverse information-gathering needs.

How It Works

The architecture comprises two layers: a CLI executor responsible for running deterministic commands, provider routing, fallbacks, output formatting, and local configuration; and a Skill/AI orchestration layer that infers user intent, selects between live search or Deep Research, executes planned CLI steps, and synthesizes source-backed answers. Its core approach leverages a skill system for AI integration and differentiates between immediate search queries and the explicit offline planning capabilities of deep research, which generates a structured research_plan for step-by-step execution.

Quick Start & Requirements

• Primary install: npm install -g @konbakuyomu/smart-search@latest (Stable) or npm install -g @konbakuyomu/smart-search@next (Test).
• Prerequisites: Node.js / npm, Python 3.10 or newer (available as python, python3, or py -3).
• Setup: Run smart-search setup for interactive configuration or smart-search setup --non-interactive for scripted setup.
• Configuration: Providers can be configured via smart-search setup or environment variables.

Highlighted Details

• CLI-first, skill-driven web research framework for AI agents and terminal users.
• Reproducible command layer supporting live search, source discovery, page fetching, site mapping, diagnostics, and offline Deep Research planning.
• Dual modes: fast live search and offline deep research planning generating actionable research_plans.
• Extensive provider support (xAI, OpenAI-compatible, Exa, Context7, Zhipu, Tavily, Firecrawl) with automatic fallback chains.
• Outputs results in JSON, Markdown, or compact content formats, facilitating integration and human readability.

Maintenance & Community

The README does not detail specific contributors, sponsorships, or community channels like Discord or Slack. It acknowledges the "LinuxDo community for the discussions that shaped the CLI + Skills workflow."

Licensing & Compatibility

• License: MIT.
• Compatibility: The MIT license permits commercial use and integration within closed-source projects without significant restrictions.

Limitations & Caveats

The deep research planner operates offline by default, requiring explicit execution of its generated plan for live research. The tool relies on users providing API keys for various third-party services. The npm package installs an isolated Python runtime, which may add complexity to environment management.