0:00 / 0:00
News
Claude Code Routines Are Here - Scheduled Agents, GitHub Triggers and API Access on Anthropic's Cloud
calendar_today Date:
database
Summary Report
Anthropic launches Claude Code Routines - templated agents that run scheduled, trigger from GitHub events, or via API on Anthropic's cloud infrastructure.
- 01. Claude Code Routines combine prompts, repositories, and connectors into templated agents that run on Anthropic's cloud
- 02. Three trigger types available: scheduled execution, GitHub events, and API calls via HTTP POST
- 03. Anthropic uses Routines internally for documentation and backlog maintenance
- 04. API triggers transform Claude Code from developer tool into programmable infrastructure
- 05. Routines enable autonomous agent operation without requiring user laptops to remain active
Anthropic has launched Routines for Claude Code, introducing templated agents that operate autonomously on the company's cloud infrastructure. These Routines combine a prompt, repositories, and MCP connectors such as Slack, Linear, Google Drive, or GitHub into a single configuration that runs without user intervention.
Routines support three trigger types: scheduled execution, GitHub events, and API calls. Scheduled routines can run hourly, nightly, on weekdays, or weekly. GitHub triggers respond to over a dozen event types including pull requests, commits, issue labels, and workflow completions. The API trigger allows external systems to invoke routines via HTTP POST requests to dedicated endpoints.
Anthropic reports that Routines have already transformed their internal documentation and backlog maintenance processes, demonstrating real-world application beyond theoretical use cases. The API functionality particularly stands out, as it transforms Claude Code from a developer tool into programmable infrastructure that other systems can invoke when computational work is required.
This development represents a significant shift towards Claude Code operating as a background service, potentially changing how developers integrate AI agents into their workflows and systems architecture.
Related Stories
play_arrow
Codex Just Got a Massive Update - Mac Apps, Memory, Automations and Image Generation
play_arrow
Anthropic Releases Claude Opus 4.7 - Built for Long-Running Tasks With Self-Verification
play_arrow
GPT-5.4-Cyber Is Here - OpenAI's Answer to Claude Mythos in the AI Cybersecurity Race
play_arrow
Google Chrome Just Added AI Skills - Save a Prompt Once and Run It Anywhere With One Click