Conversation
Co-authored-by: pelikhan <4175913+pelikhan@users.noreply.github.com>
| issues: | ||
| types: [opened] | ||
| permissions: | ||
| issues: write |
There was a problem hiding this comment.
@copilot permissions should be read-all . Write is done in the safe outputs
There was a problem hiding this comment.
Updated the permissions to use read-all in the example (commit 78be9e9). Write operations are now clearly handled through safe-outputs as intended.
Co-authored-by: pelikhan <4175913+pelikhan@users.noreply.github.com>
Added a "Security Benefits" section explaining the three key security features: read-only permissions by default, sanitized safe-outputs for write operations, and team member access gating (commit 78be9e9). Warning Firewall rules blocked me from connecting to one or more addresses (expand for details)I tried to connect to the following addresses, but was blocked by firewall rules:
If you need me to access, download, or install something from one of these locations, you can either:
|
|
@dsyme a small update to the readme to make it a bit more real |
|
I get what you're doing here, though somehow I think any README-based landing page in a githubnext repo just isn't going to cut it. We need a proper landing page, and best to get people to it as soon as possible rather than hanging around in the repo |
|
I can get a Astro starlight documentation site going as well on the side but I wanted to show the simplicity here. Probably make this sample even shorter. |
…5460-4532-bfaa-fad91d688f26
…gent functionality
Enhances the README to make it clearer what GitHub Agentic Workflows is about by adding a new "How It Works" section with a concrete example, visual diagram, and security information.
What's Changed
The README now includes a practical demonstration of the core concept: transforming natural language markdown into executable GitHub Actions workflows. The new section shows:
Concrete Example: A simple issue auto-responder workflow that demonstrates:
Visual Flow Diagram: A mermaid diagram illustrating the complete workflow:
Clear Explanation: Plain English summary explaining that "the AI agent reads your repository context, understands the issue content, and takes appropriate actions - all defined in natural language rather than complex code."
Security Benefits: Brief explanation of the security model including read-only permissions by default, sanitized safe-outputs for write operations, and team member access gating.
Why This Helps
The original README jumped straight to documentation links without showing what the tool actually does. New users can now immediately see:
The example chosen (issue auto-responder) is simple enough for newcomers but demonstrates key concepts like event triggers, permissions, and safe outputs that make agentic workflows powerful and secure.
This addresses the issue request for "a short example of agentic workflow and a mermaid diagram of the generated workflow" with minimal changes that enhance clarity without disrupting the existing content structure.
💡 You can make Copilot smarter by setting up custom instructions, customizing its development environment and configuring Model Context Protocol (MCP) servers. Learn more Copilot coding agent tips in the docs.