Skip to content

[workflow-style] Normalize report formatting for auto-triage-issues #11275

Description

@github-actions

Workflow to Update

Workflow File: .github/workflows/auto-triage-issues.md

Issue: This workflow automatically triages issues every 6 hours and creates triage reports, but doesn't include markdown style guidelines. Triage reports with analysis and recommendations would benefit from consistent formatting.

Why This Matters

The auto-triage-issues workflow runs frequently (every 6h) and creates comments/reports about issue classification, priority, and routing. These triage reports should be scannable yet comprehensive, making them ideal for progressive disclosure.

Required Changes

Update the workflow prompt to include these formatting guidelines (add after the triage objectives):

1. Header Levels

Add instruction: "Use h3 (###) or lower for all headers in triage reports to maintain proper document hierarchy."

Rationale: Triage reports contain classification, severity assessment, routing recommendations, and supporting analysis. Proper hierarchy keeps these sections organized.

2. Progressive Disclosure

Add instruction: "Wrap detailed analysis and supporting evidence in <details><summary><b>Section Name</b></summary> tags to improve readability."

Example:

<details>
<summary><b>View Detailed Classification Analysis</b></summary>

[Detailed reasoning for classification, keyword analysis, pattern matching results]

</details>

Rationale: Triage decisions should show the conclusion clearly (label, priority, assignee), but detailed reasoning can be hidden to avoid cluttering issue threads.

3. Report Structure

Suggest adding this structure pattern:

### Triage Summary
- **Classification**: [bug/feature/question/etc]
- **Priority**: [P0/P1/P2/P3]
- **Suggested Labels**: [list]
- **Suggested Assignee**: `@username` or team

### Routing Recommendation
[Clear, actionable recommendation - always visible]

<details>
<summary><b>View Classification Details</b></summary>

[Why this classification was chosen, confidence score, keywords detected]

</details>

<details>
<summary><b>View Similar Issues</b></summary>

[Links to similar issues, patterns detected, historical context]

</details>

### Confidence Assessment
- **Overall Confidence**: [High/Medium/Low]
- **Reasoning**: [Brief explanation - keep visible]

Design Principles (Airbnb-Inspired)

The updated workflow should create reports that:

  1. Build trust through clarity: Triage decision and routing immediately visible
  2. Exceed expectations: Include confidence scores, similar issues, detailed reasoning
  3. Create delight: Use progressive disclosure to share thorough analysis without cluttering
  4. Maintain consistency: Follow the same patterns as other automated classification workflows

Import Statement to Add

Add this to the imports: section in the frontmatter:

imports:
  - shared/reporting.md

This automatically loads the formatting guidelines into the agent's context.

Example Reference

See workflows like issue-classifier.md or issue-triage-agent.md for examples of classification workflows (once they're updated with reporting guidelines).

Agent Task

  1. Add shared/reporting.md to the imports section
  2. Add formatting guidelines to the prompt instructions (emphasize triage comment structure)
  3. Test by letting the workflow triage a few issues and reviewing the comment format
  4. Ensure triage decisions stay visible while detailed analysis is collapsible
  5. Verify comments don't clutter issue threads with excessive visible detail

AI generated by Workflow Normalizer

  • expires on Jan 29, 2026, 12:27 PM UTC

Metadata

Metadata

Labels

cookieIssue Monster Loves Cookies!documentationImprovements or additions to documentationenhancementNew feature or requestworkflow-style

Type

No type
No fields configured for issues without a type.

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions