Plugin Issue Triage

Dart & Flutter IntelliJ Plugin Issue Triage Skill

This skill guides the process of fetching, analyzing, reviewing, and applying triage decisions to GitHub issues in the flutter/dart-intellij-third-party and flutter/flutter-intellij repositories using a local interactive dashboard.

[!IMPORTANT] WRITING GUIDELINES When drafting replies, explanations, or any prose, refer to the natural-writing skill to ensure clarity, accuracy, and tone.

Workflow

When the user invokes this skill, first ask them if they want to triage the dart plugin (flutter/dart-intellij-third-party), the flutter plugin (flutter/flutter-intellij), or both. If they select both, perform the following steps across both repositories and merge the results into a single combined view.

Step 1: Fetch Untriaged Issues

Use the fetch script to retrieve all open issues lacking a priority label in the selected repository (or both).

1. Run the fetch script to download the issues to a raw JSON file in your conversation-specific scratch directory: python3 .agents/skills/plugin-issue-triage/scripts/fetch_issues.py --repo "<REPO_NAME>" --output-file "<appDataDir>/brain/<conversation-id>/scratch/raw_issues_<REPO_NAME_CLEANED>.json" (Replace <REPO_NAME> with flutter/dart-intellij-third-party or flutter/flutter-intellij, and <REPO_NAME_CLEANED> with a filesystem-safe version like dart or flutter). If "both" was selected, run this script twice and keep both JSON files.

Step 2: Analyze and Suggest Triage

Process the raw issues and generate recommended triage fields based on the project's triage criteria.

1. Read the triage criteria reference document: triage_criteria.md, priorities.md, and labels.md.
2. Natively Orchestrate Subagents:
   • Load the first N issues (defaulting to 30, or as requested) from the downloaded JSON. If "both" was selected, load up to 30 issues from each repository.
   • Call the invoke_subagent tool in parallel for those issues. Prompt each subagent to analyze its assigned issue against the guidelines in triage_criteria.md, proposed_actions.md, priorities.md, and labels.md. Instruct them to return a structured JSON block containing issue_type ("bug", "feature", or "task"), priority, proposed_actions (an array of tag strings from proposed_actions.md), labels, reply, and search_keywords (a string of 3-5 broad, distinctive keywords, OR a specific stack trace snippet, designed to find duplicate issues without being overly restrictive).
   • While the subagents are running, or after they report back, use the gh issue list -R <repo> --search "<search_keywords>" --state all --json number,title,state,createdAt,url --limit 3 command for each issue's generated keywords.
   • Compile their recommendations and your search results into the standard schema:
     • Inject the subagent's JSON recommendations into the issue object under the key suggestions.
     • Ensure assignee is left empty by default unless there is a strong reason to assign an owner.
     • Inject possible_duplicates (the raw JSON array of the top 3 GitHub search results, omitting the current issue itself) into the suggestions object.
     • Inject total_issues_count (preserving the total count from the raw issues JSON) at the root level of the JSON.
     • Critically, ensure that the root-level metadata structures (labels_by_repo, owners, and assignees) from the raw JSON payload(s) are strictly preserved and merged into the compiled JSON so the dashboard has access to them.
     • Save the final compiled payload to issues_to_triage_<REPO_NAME_CLEANED>.json in the scratch directory. If "both" was selected, combine all compiled issues into a single issues_to_triage_combined.json file. Ensure the combined issues are grouped by repository (all dart-intellij-third-party issues first, followed by all flutter-intellij issues) rather than interleaved.

Step 3: Launch Review Dashboard

Launch the interactive web dashboard to allow the engineer to review and refine the suggested triages.

1. Start the local server as a background task, pointing it to your scratch directory: python3 .agents/skills/plugin-issue-triage/scripts/launch_dashboard.py --data-file "<appDataDir>/brain/<conversation-id>/scratch/issues_to_triage_<REPO_NAME_CLEANED>.json" --output-file "<appDataDir>/brain/<conversation-id>/scratch/triage_decisions_<REPO_NAME_CLEANED>.json" (If "both" was selected, use issues_to_triage_combined.json and triage_decisions_combined.json) Set WaitMsBeforeAsync to 1000 so the server runs in the background.
2. Wait for Completion: Stop calling tools and go idle. The launcher will automatically open the browser for the user and block until they click "Apply Triages" or "Abort". Once the user acts, the background task will complete, and you will receive a notification with the exit status.
3. Verify Exit Status:
   • If the task exited with status 0 (approved), proceed to apply the decisions.
   • If the task exited with a non-zero status (abort), don't modify any issues. You MUST STOP and ask the user for further instructions.

Step 4: Apply Decisions to GitHub

Once the dashboard task exits successfully, Execute Approved Decisions: Run the apply script to update the approved labels, assignees, and comments on GitHub:

python3 .agents/skills/plugin-issue-triage/scripts/apply_triage.py --decisions-file "<appDataDir>/brain/<conversation-id>/scratch/triage_decisions_<REPO_NAME_CLEANED>.json" (If "both" was selected, use triage_decisions_combined.json)

Bundled Resources

• triage_criteria.md: Guide for classifying issues and assigning priorities.
• priorities.md: Priority mappings and guidelines.
• labels.md: Label heuristics and guidelines.
• scripts/fetch_issues.py: Script to query open, untriaged issues via the GitHub CLI.
• scripts/launch_dashboard.py: Local HTTP server that opens the interactive web dashboard in the user's browser.
• assets/triage_dashboard.html: HTML/CSS/JS template for the triage review interface.
• scripts/apply_triage.py: Script to apply finalized triage decisions to GitHub.