This is an early release preview. You may encounter bugs.
Tools logo
Unclaimed

Tool coding connectors

Tools

A 88/100 GitHub score ? This grade is derived from GitHub signals, not user votes. Open for the full breakdown.
Discussion

01 / Summary

About Tools.

Strands Agents Tools is a Python package of ready-made tools for agents built with the Strands Agents SDK. Each tool is a function an agent can call: reading and writing files, running shell commands and Python code, making HTTP requests, searching the web, storing memories, driving a browser or desktop, generating images and video, and coordinating other agents.

Tools are added to an agent's tool list and invoked either by the model during a run or directly from code. Several tools wrap external services: Mem0, Amazon Bedrock Knowledge Bases, Elasticsearch, and MongoDB Atlas for memory, Tavily and Exa for search and crawling, TwelveLabs for video search, and Amazon Bedrock Nova Reel for video generation. Others expose the agent to protocols, such as an A2A client for talking to A2A-compliant agents and a dynamic MCP client that loads tools from remote MCP servers at runtime.

The tools are marked experimental, and many are now deprecated as their capabilities move into the core SDK (vended file editor, shell, and sleep tools, concurrent tool execution, native extended thinking, a memory manager) or into vendor-maintained MCP servers. Deprecated tools keep working but log a warning, and the package is expected to be archived once the migration is complete.

  • File operations: read, write, and edit files, including pattern replacement and multi-file edits
  • Code and shell execution: run shell commands and Python snippets with state persistence and user confirmation before execution
  • Memory backends: store and retrieve user and agent memories with Mem0, Amazon Bedrock Knowledge Bases, AgentCore Memory, Elasticsearch, or MongoDB Atlas
  • Web search and crawling: search, extract page content, crawl, and map sites through Tavily and Exa tools
  • Agent as tool: spawn nested agents with their own system prompt, tools, and model provider for specialized sub-tasks
  • Multi-agent patterns: swarm coordination with shared memory and DAG-based multi-agent graphs with per-node model configuration
  • Dynamic MCP client: connect to external MCP servers at runtime and load their tools into the agent
  • A2A client: discover A2A-compliant agents and send messages between them
  • Browser and computer control: automate a Chromium browser and desktop actions such as mouse, keyboard, and screenshots
  • Media tools: image generation, video generation with Nova Reel, video search and chat through TwelveLabs models, and audio output
  • Utilities: calculator with symbolic math, cron scheduling, RSS feed management, diagram generation, journaling, and Slack access

02 / Discussion CREDIBILITY-GATED

Discussion

Reading is open to everyone. Posting and voting need a verified identity or a GitHub grade of B or higher.

  • No discussions yet.

04 / Build

Build with Tools.

Browse the catalog for frameworks, tools, and harnesses, each scored on real GitHub credibility.

Get Tools →

Browse the catalog