---
name: Ahk Triage
slug: ahk-triage
category: Quality
description: Ahk Triage analyzes a bug or unexpected behavior and produces a structured diagnostic report. Use it when you need root-cause investigation without creating tasks or tracking harness work.
github: "https://github.com/enmanuelmag/agent-harness-kit/tree/main/.agents/skills/ahk-triage"
language: TypeScript
stars: 180
forks: 9
install: "npx degit https://github.com/enmanuelmag/agent-harness-kit/tree/main/.agents/skills/ahk-triage ~/.claude/skills/ahk-triage"
installs_to: ~/.claude/skills/ahk-triage
source_path: .agents/skills/ahk-triage/SKILL.md
collection_size: 14
category_size: 1817
collection_url: "https://dirskills.com/collections/enmanuelmag/agent-harness-kit"
added: 2026-09-07T05:20:26.853Z
last_synced: 2026-09-07T05:20:26.853Z
canonical_url: "https://dirskills.com/skills/ahk-triage"
---

# Ahk Triage

Ahk Triage analyzes a bug or unexpected behavior and produces a structured diagnostic report. Use it when you need root-cause investigation without creating tasks or tracking harness work.

**Install:**

```bash
npx degit https://github.com/enmanuelmag/agent-harness-kit/tree/main/.agents/skills/ahk-triage ~/.claude/skills/ahk-triage
```

## README

## Provider Delegation Guidance

- Sequential: Delegate only bounded, independent work; retain decisions and the final synthesis in the parent thread.
- Parallel: Delegate independent work in parallel when useful, then wait for every delegated result before continuing.
- Context Transfer: Give every delegated task a self-contained objective, scope, relevant context, restrictions, and output contract.
- Wait For Completion: Wait for the delegated result, then consolidate its findings in the parent thread.
- Inspect Progress: In the interactive CLI, use /agent to inspect delegated threads when needed.


You are in **lightweight triage mode**.

> If `$ARGUMENTS` is empty, ask the user to describe the issue before doing anything else.

The issue: $ARGUMENTS

## Rules
- NO MCP calls — no tasks.*, no actions.*
- Run health.sh ONLY if the issue explicitly involves the build or test suite — not as a routine step
- NO builder, NO reviewer
- Create files ONLY if user explicitly asked to save the report

## Process

1. Invoke **Explorer** as a subagent:
   > "Read-only diagnostic investigation — no MCP harness. Issue reported: `$ARGUMENTS`. Investigate: which files are likely involved, what code paths produce this behavior, what assumptions might be violated, what state or environment conditions could trigger this. Return a detailed diagnostic analysis."

2. Synthesize Explorer's findings into the REQUIRED output format below. Do not skip or reorder sections.

## Required output format

---

**TL;DR**
One paragraph. What is happening and the most likely cause.

**Long Description**
Detailed analysis: what the code is doing, what it should do, where the divergence occurs.

**Root Cause Analysis**
The causal chain: trigger → behavior → symptom. Name specific files, functions, and line numbers where relevant.

**Possible Fixes**
Ordered most-to-least likely. Each fix must be a concrete actionable step naming specific files/functions.

**Follow-up Questions**
*(Omit this section entirely if investigation was sufficient. Include only genuine unknowns that would change the diagnosis.)*

---