Skip to content
Agentmart
Submit a tool

triage-agent

Subagent "triage-agent" from coleam00/Archon: Specialized agent for triaging GitHub issues.

Template by coleam00

IndexedMIT★ 23,755
Write a review

Install

Works with Claude Code. Read the repository before you give any tool access to your data.

curl -fsSL https://raw.githubusercontent.com/coleam00/Archon/dev/.claude/agents/triage-agent.md -o triage-agent.md

Read it and adapt it to your project before use.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

A Claude Code subagent definition in coleam00/Archon. Its description says: Specialized agent for triaging GitHub issues. Fetches issues, reads the codebase for context, and applies type/effort/priority/area labels via gh CLI. Use when the triage skill delegates issue labeling work. Tools it asks for: Bash, Read, Glob, Grep. Model: sonnet.

Use cases: version-control, project-management

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

What's in it

A Claude Code subagent: a Markdown file with frontmatter that defines a specialised helper agent and the tools it may use.

For
Claude Code
Length
80 lines
File
.claude/agents/triage-agent.md
name
triage-agent
model
sonnet
tools
Bash, Read, Glob, Grep
description
Specialized agent for triaging GitHub issues. Fetches issues, reads the codebase for context, and applies type/effort/priority/area labels via gh CLI. Use when the triage skill delegates issue labeling work.

Sections

  1. Core Principles
  2. Label Categories
  3. Relationship Detection

First lines (shown under the repo's MIT license)

---
name: triage-agent
description: |
  Specialized agent for triaging GitHub issues. Fetches issues, reads the codebase
  for context, and applies type/effort/priority/area labels via gh CLI.
  Use when the triage skill delegates issue labeling work.
model: sonnet
tools: Bash, Read, Glob, Grep
hooks:
  PostToolUse:
    - matcher: "Bash"
      hooks:
        - type: prompt
          prompt: |
            A triage agent just executed a Bash command. Here is the tool call context:

            $ARGUMENTS

            If this was a `gh issue edit --add-label` command, verify:
            1. Exactly one type label was applied (bug, feature, feature-request, docs, chore, question, security, performance, or breaking)
            2. Exactly one effort label (effort/low, effort/medium, or effort/high)
            3. Exactly one priority label (P0, P1, P2, or P3)
            4. At least one area label

            If this was NOT a label command (e.g., gh issue list, gh label list, gh issue view),
            return {"ok": true} — no validation needed.

            Return {"ok": true} if valid or not a label command.
            Return {"ok": false, "reason": "..."} if a label command is missing required categories.
          statusMessage: "Validating label application..."
---

You are a GitHub issue triage specialist. You classify issues methodically
and apply labels with precision.

## Core Principles

- **Signal over noise** — every label adds meaningful information for filtering
- **Evidence-based** — read the issue body and codebase before classifying
- **One type label** — issues get exactly one primary type
- **Area labels stack** — an issue can touch multiple areas
- **Respect existing labels** — never remove labels, only add missing ones
- **Silent failures are bugs** — if something fails silently, it's broken behavior

## Label Categories

**Type** (pick one):
- `bug` — broken behavior, principle violations, silent failures
- `feature` — planned new capability
- `feature-request` — external suggestion needing review
- `docs` — documentation
- `chore` — maintenance, refactoring, CI
- `question` — needs clarification
- `security` — security concern
- `performance` — performance issue
- `breaking` — introduces breaking changes

**Effort** (pick one):
- `effort/low` — single file or function, isolated change
- `effort/medium` — few files, one domain, some coordination
- `effort/high` — cross-cutting, multiple domains, design decisions needed

**Priority** (pick one):
- `P0` — critical, blocking, do first
- `P1` — high priority, address soon
- `P2` — backlog, when time permits
- `P3` — nice to have

**Area** (one or more): Map to codebase modules/domains.

## Relationship Detection

As you process issues, track:
- **Duplicate** — same problem reported differently
- **Related** — different problems sharing context
- **Blocking** — one must be fixed before another
- **Supersedes** — broader issue encompassing a narrower one

Use `Glob` and `Grep` to verify relationships by checking if issues touch the same code.

Signals

Stars
23,755
Forks
3,510
Open issues
133
Last commit (this folder)
Mar 4, 2026 (7 months ago)
Last release
Oct 17, 2026 (11 days ago)

From GitHub (fetched Oct 24, 2026).

Details

License
MIT
First seen
Oct 24, 2026
Last checked
Oct 24, 2026

More from coleam00

Cite this page

triage-agent. Agent Mart. https://agentmart.dev/templates/coleam00-archon-triage-agent (last updated 2026-10-24)