issue-triage-loop
Skill6 starsChanged 4 months agoMITView on GitHub
Batch-process issues through improve → triage → retitle, one subagent per issue, sequentially. Invoke with the issue tracker CLI, project identifier, and issue selection criteria.
In this repository14 other files it gives its agents.
- strict-code-reviewSkill
- fabro-workflowSkill
- fastapi-proSkill
- industry-standardizeSkill
- issue-decomposerSkill
- martin-clean-codeSkill
- pr-code-reviewSkill
- prd-reviewSkill
- rebase-on-mainSkill
- request-to-prdSkill
- startup-analystSkill
- two-axis-reviewSkill
- un-rgeSkill
- unslopSkill
The file
We split this skill into the rules it gives an agent
3 rules. About 0 are common, 3 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈3
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 20
Ask for any missing piece before continuing.
- Distinctive: line 36
Wait for it to finish before continuing.
- Distinctive: line 72
After all issues: report total processed, any failures with issue IDs and error details.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Issue Triage Loop
A relay: each issue runs its own subagent leg — improve, triage, retitle — one at a time, so the main context never accumulates per-issue state.
Orient
Establish before starting:
- Tracker CLI — tool and invocation pattern (e.g.
ghfor GitHub,glabfor GitLab, or any user-provided CLI) - Project — repo or project reference (e.g.
owner/repofor GitHub,group/projectfor GitLab) - Commands — how to: fetch an issue with its body and labels; update its body; update its labels; post a comment; update its title
- Selection — which issues to process: a numeric range, a label filter, "all open", or any expression the CLI supports
Ask for any missing piece before continuing. Do not guess CLI syntax — wrong commands silently corrupt issues.
Enumerate
In the main context, run the CLI to fetch every issue ID matching the selection. Produce an ordered list.
Confirm the count with the user before starting the relay.
Completion criterion: every matching issue ID is in the list and the user has confirmed.
Relay
For each issue ID in order:
- Fill in the relay brief below with this issue's CLI context.
- Spawn one subagent with that brief.
- Wait for it to finish before continuing.
- Log the result — issue ID, outcome, new title — before moving to the next.
Relay brief template
Fill every {PLACEHOLDER} before sending.
You are processing issue {ISSUE_ID} in {PROJECT} using {CLI_TOOL}.
CLI commands for this issue:
Fetch: {GET_COMMAND}
Update body: {UPDATE_BODY_COMMAND}
Update labels: {UPDATE_LABELS_COMMAND}
Add comment: {ADD_COMMENT_COMMAND}
Update title: {UPDATE_TITLE_COMMAND}
Fetch the issue now (body, title, current labels). Then run these three steps in order. Complete each step fully before starting the next.
Step 1 — Improve
Invoke the issue-improver skill on the fetched issue body. Apply its output by updating the issue body in place using the update-body command.
Done when: the issue body on the tracker matches the improver's output.
Step 2 — Triage
Invoke the issue-triage skill on the issue. Apply its output: update the issue's labels using the update-labels command; post the triage output as a comment using the add-comment command.
Done when: labels are updated and the triage comment is posted on the tracker.
Step 3 — Retitle
Read the current issue (improved body, triage comment). Write a new title as a conventional commits message: type(scope): description. Update the title using the update-title command.
Done when: the title on the tracker is a valid conventional commits message.
Return one line: issue ID | old title | new title | labels appliedWrap up
After all issues: report total processed, any failures with issue IDs and error details.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

