Skip to content
Closed
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
49 changes: 49 additions & 0 deletions AGENTS.md
Original file line number Diff line number Diff line change
Expand Up @@ -988,6 +988,55 @@ Before EVERY commit:
- Use conventional commits for commit messages
- do NOT commit explanation markdown files about the fixes

## Workflow Quality Benchmarks

Based on [Agent Persona Exploration research](https://githubnext.github.io/gh-aw/research/agent-persona-exploration-2026-01/), agentic workflows should meet these quality standards:

### Production-Ready Quality Score: 4.97/5.0

Research across 6 diverse scenarios achieved an average quality score of 4.97/5.0, establishing a high benchmark for workflow quality.

### Quality Score Framework

| Score | Quality Level | Characteristics |
|-------|--------------|-----------------|
| 5.0 | Exceptional | Production-ready, comprehensive docs, best practices throughout |
| 4.5-4.9 | Excellent | Production-ready, good docs, minor improvements possible |
| 4.0-4.4 | Good | Functional, basic docs, some refinement needed |
| 3.5-3.9 | Adequate | Works but needs improvement, limited docs |
| < 3.5 | Needs Work | Significant issues, incomplete, or poor quality |

### Expected Quality Standards

All production workflows should demonstrate:

- ✅ **Clear purpose**: Single, well-defined responsibility
- ✅ **Minimal permissions**: Read-only with safe-outputs for writes (100% compliance in research)
- ✅ **Appropriate triggers**: Matches use case requirements
- `pull_request` for code review and validation (50% of workflows)
- `schedule` for periodic analysis (33% of workflows)
- `workflow_run` for post-deployment actions (17% of workflows)
- ✅ **Proper tools**: Only what's needed, properly configured
- GitHub tools: Universal (100% of workflows)
- Playwright: For browser automation when needed
- AI analysis: For pattern recognition and insights
- ✅ **Complete documentation**: At least README with setup and examples
- ✅ **Error handling**: Graceful failures with actionable messages
- ✅ **Business value**: Clear ROI or benefit statement

### Security Compliance (100% in Research)

All workflows must maintain:
- **Read-only permissions by default**: Never request write permissions in frontmatter
- **Safe-outputs pattern**: All write operations through sanitized safe-outputs
- **Network isolation**: Restrict access to required domains only
- **No secret exposure**: Automatic sanitization of sensitive data

For detailed patterns and best practices, see:
- [Workflow Patterns & Best Practices](https://githubnext.github.io/gh-aw/best-practices/workflow-patterns/)
- [Agent Persona Exploration Research](https://githubnext.github.io/gh-aw/research/agent-persona-exploration-2026-01/)
- [Security Best Practices](https://githubnext.github.io/gh-aw/guides/security/)

## Operational Runbooks

For investigating and resolving workflow issues:
Expand Down
2 changes: 2 additions & 0 deletions README.md
Original file line number Diff line number Diff line change
Expand Up @@ -61,6 +61,8 @@ Security is foundational to GitHub Agentic Workflows. Workflows run with read-on

For complete documentation, examples, and guides, see the [Documentation](https://githubnext.github.io/gh-aw/).

**Research & Best Practices**: Learn from our [Agent Persona Exploration research](https://githubnext.github.io/gh-aw/research/agent-persona-exploration-2026-01/) which analyzed 6 production scenarios achieving a 4.97/5.0 quality score, and explore proven [workflow patterns](https://githubnext.github.io/gh-aw/best-practices/workflow-patterns/) for triggers, tools, security, and documentation.

## Contributing

We welcome contributions to GitHub Agentic Workflows! Here's how you can help:
Expand Down
12 changes: 12 additions & 0 deletions docs/astro.config.mjs
Original file line number Diff line number Diff line change
Expand Up @@ -174,6 +174,12 @@ export default defineConfig({
{ label: 'Ephemerals', link: '/guides/ephemerals/' },
],
},
{
label: 'Best Practices',
items: [
{ label: 'Workflow Patterns', link: '/best-practices/workflow-patterns/' },
],
},
{
label: 'Design Patterns',
items: [
Expand Down Expand Up @@ -241,6 +247,12 @@ export default defineConfig({
label: 'Troubleshooting',
autogenerate: { directory: 'troubleshooting' },
},
{
label: 'Research',
items: [
{ label: 'Agent Persona Exploration (2026-01)', link: '/research/agent-persona-exploration-2026-01/' },
],
},
],
}),
],
Expand Down
1 change: 1 addition & 0 deletions docs/src/content/docs/agent-factory-status.mdx
Original file line number Diff line number Diff line change
Expand Up @@ -75,6 +75,7 @@ These are experimental agentic workflows used by the GitHub Next team to learn,
| [Discussion Task Miner - Code Quality Improvement Agent](https://github.com/githubnext/gh-aw/blob/main/.github/workflows/discussion-task-miner.md) | copilot | [![Discussion Task Miner - Code Quality Improvement Agent](https://github.com/githubnext/gh-aw/actions/workflows/discussion-task-miner.lock.yml/badge.svg)](https://github.com/githubnext/gh-aw/actions/workflows/discussion-task-miner.lock.yml) | - | - |
| [Documentation Noob Tester](https://github.com/githubnext/gh-aw/blob/main/.github/workflows/docs-noob-tester.md) | copilot | [![Documentation Noob Tester](https://github.com/githubnext/gh-aw/actions/workflows/docs-noob-tester.lock.yml/badge.svg)](https://github.com/githubnext/gh-aw/actions/workflows/docs-noob-tester.lock.yml) | - | - |
| [Documentation Unbloat](https://github.com/githubnext/gh-aw/blob/main/.github/workflows/unbloat-docs.md) | claude | [![Documentation Unbloat](https://github.com/githubnext/gh-aw/actions/workflows/unbloat-docs.lock.yml/badge.svg)](https://github.com/githubnext/gh-aw/actions/workflows/unbloat-docs.lock.yml) | - | `/unbloat` |
| [Draft PR Cleanup](https://github.com/githubnext/gh-aw/blob/main/.github/workflows/draft-pr-cleanup.md) | copilot | [![Draft PR Cleanup](https://github.com/githubnext/gh-aw/actions/workflows/draft-pr-cleanup.lock.yml/badge.svg)](https://github.com/githubnext/gh-aw/actions/workflows/draft-pr-cleanup.lock.yml) | - | - |
| [Duplicate Code Detector](https://github.com/githubnext/gh-aw/blob/main/.github/workflows/duplicate-code-detector.md) | codex | [![Duplicate Code Detector](https://github.com/githubnext/gh-aw/actions/workflows/duplicate-code-detector.lock.yml/badge.svg)](https://github.com/githubnext/gh-aw/actions/workflows/duplicate-code-detector.lock.yml) | - | - |
| [Example: Custom Error Patterns](https://github.com/githubnext/gh-aw/blob/main/.github/workflows/example-custom-error-patterns.md) | copilot | [![Example: Custom Error Patterns](https://github.com/githubnext/gh-aw/actions/workflows/example-custom-error-patterns.lock.yml/badge.svg)](https://github.com/githubnext/gh-aw/actions/workflows/example-custom-error-patterns.lock.yml) | - | - |
| [Example: Properly Provisioned Permissions](https://github.com/githubnext/gh-aw/blob/main/.github/workflows/example-permissions-warning.md) | copilot | [![Example: Properly Provisioned Permissions](https://github.com/githubnext/gh-aw/actions/workflows/example-permissions-warning.lock.yml/badge.svg)](https://github.com/githubnext/gh-aw/actions/workflows/example-permissions-warning.lock.yml) | - | - |
Expand Down
Loading